discontinued / server · Files

REC-PNG-20031110 311 KB
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746 4747 4748 4749 4750 4751 4752 4753 4754 4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773 4774 4775 4776 4777 4778 4779 4780 4781 4782 4783 4784 4785 4786 4787 4788 4789 4790 4791 4792 4793 4794 4795 4796 4797 4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821 4822 4823 4824 4825 4826 4827 4828 4829 4830 4831 4832 4833 4834 4835 4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846 4847 4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 4858 4859 4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871 4872 4873 4874 4875 4876 4877 4878 4879 4880 4881 4882 4883 4884 4885 4886 4887 4888 4889 4890 4891 4892 4893 4894 4895 4896 4897 4898 4899 4900 4901 4902 4903 4904 4905 4906 4907 4908 4909 4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996 4997 4998 4999 5000 5001 5002 5003 5004 5005 5006 5007 5008 5009 5010 5011 5012 5013 5014 5015 5016 5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050 5051 5052 5053 5054 5055 5056 5057 5058 5059 5060 5061 5062 5063 5064 5065 5066 5067 5068 5069 5070 5071 5072 5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118 5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138 5139 5140 5141 5142 5143 5144 5145 5146 5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186 5187 5188 5189 5190 5191 5192 5193 5194 5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213 5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263 5264 5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283 5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300 5301 5302 5303 5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320 5321 5322 5323 5324 5325 5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336 5337 5338 5339 5340 5341 5342 5343 5344 5345 5346 5347 5348 5349 5350 5351 5352 5353 5354 5355 5356 5357 5358 5359 5360 5361 5362 5363 5364 5365 5366 5367 5368 5369 5370 5371 5372 5373 5374 5375 5376 5377 5378 5379 5380 5381 5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392 5393 5394 5395 5396 5397 5398 5399 5400 5401 5402 5403 5404 5405 5406 5407 5408 5409 5410 5411 5412 5413 5414 5415 5416 5417 5418 5419 5420 5421 5422 5423 5424 5425 5426 5427 5428 5429 5430 5431 5432 5433 5434 5435 5436 5437 5438 5439 5440 5441 5442 5443 5444 5445 5446 5447 5448 5449 5450 5451 5452 5453 5454 5455 5456 5457 5458 5459 5460 5461 5462 5463 5464 5465 5466 5467 5468 5469 5470 5471 5472 5473 5474 5475 5476 5477 5478 5479 5480 5481 5482 5483 5484 5485 5486 5487 5488 5489 5490 5491 5492 5493 5494 5495 5496 5497 5498 5499 5500 5501 5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 5519 5520 5521 5522 5523 5524 5525 5526 5527 5528 5529 5530 5531 5532 5533 5534 5535 5536 5537 5538 5539 5540 5541 5542 5543 5544 5545 5546 5547 5548 5549 5550 5551 5552 5553 5554 5555 5556 5557 5558 5559 5560 5561 5562 5563 5564 5565 5566 5567 5568 5569 5570 5571 5572 5573 5574 5575 5576 5577 5578 5579 5580 5581 5582 5583 5584 5585 5586 5587 5588 5589 5590 5591 5592 5593 5594 5595 5596 5597 5598 5599 5600 5601 5602 5603 5604 5605 5606 5607 5608 5609 5610 5611 5612 5613 5614 5615 5616 5617 5618 5619 5620 5621 5622 5623 5624 5625 5626 5627 5628 5629 5630 5631 5632 5633 5634 5635 5636 5637 5638 5639 5640 5641 5642 5643 5644 5645 5646 5647 5648 5649 5650 5651 5652 5653 5654 5655 5656 5657 5658 5659 5660 5661 5662 5663 5664 5665 5666 5667 5668 5669 5670 5671 5672 5673 5674 5675 5676 5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687 5688 5689 5690 5691 5692 5693 5694 5695 5696 5697 5698 5699 5700 5701 5702 5703 5704 5705 5706 5707 5708 5709 5710 5711 5712 5713 5714 5715 5716 5717 5718 5719 5720 5721 5722 5723 5724 5725 5726 5727 5728 5729 5730 5731 5732 5733 5734 5735 5736 5737 5738 5739 5740 5741 5742 5743 5744 5745 5746 5747 5748 5749 5750 5751 5752 5753 5754 5755 5756 5757 5758 5759 5760 5761 5762 5763 5764 5765 5766 5767 5768 5769 5770 5771 5772 5773 5774 5775 5776 5777 5778 5779 5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791 5792 5793 5794 5795 5796 5797 5798 5799 5800 5801 5802 5803 5804 5805 5806 5807 5808 5809 5810 5811 5812 5813 5814 5815 5816 5817 5818 5819 5820 5821 5822 5823 5824 5825 5826 5827 5828 5829 5830 5831 5832 5833 5834 5835 5836 5837 5838 5839 5840 5841 5842 5843 5844 5845 5846 5847 5848 5849 5850 5851 5852 5853 5854 5855 5856 5857 5858 5859 5860 5861 5862 5863 5864 5865 5866 5867 5868 5869 5870 5871 5872 5873 5874 5875 5876 5877 5878 5879 5880 5881 5882 5883 5884 5885 5886 5887 5888 5889 5890 5891 5892 5893 5894 5895 5896 5897 5898 5899 5900 5901 5902 5903 5904 5905 5906 5907 5908 5909 5910 5911 5912 5913 5914 5915 5916 5917 5918 5919 5920 5921 5922 5923 5924 5925 5926 5927 5928 5929 5930 5931 5932 5933 5934 5935 5936 5937 5938 5939 5940 5941 5942 5943 5944 5945 5946 5947 5948 5949 5950 5951 5952 5953 5954 5955 5956 5957 5958 5959 5960 5961 5962 5963 5964 5965 5966 5967 5968 5969 5970 5971 5972 5973 5974 5975 5976 5977 5978 5979 5980 5981 5982 5983 5984 5985 5986 5987 5988 5989 5990 5991 5992 5993 5994 5995 5996 5997 5998 5999 6000 6001 6002 6003 6004 6005 6006 6007 6008 6009 6010 6011 6012 6013 6014 6015 6016 6017 6018 6019 6020 6021 6022 6023 6024 6025 6026 6027 6028 6029 6030 6031 6032 6033 6034 6035 6036 6037 6038 6039 6040 6041 6042 6043 6044 6045 6046 6047 6048 6049 6050 6051 6052 6053 6054 6055 6056 6057 6058 6059 6060 6061 6062 6063 6064 6065 6066 6067 6068 6069 6070 6071 6072 6073 6074 6075 6076 6077 6078 6079 6080 6081 6082 6083 6084 6085 6086 6087 6088 6089 6090 6091 6092 6093 6094 6095 6096 6097 6098 6099 6100 6101 6102 6103 6104 6105 6106 6107 6108 6109 6110 6111 6112 6113 6114 6115 6116 6117 6118 6119 6120 6121 6122 6123 6124 6125 6126 6127 6128 6129 6130 6131 6132 6133 6134 6135 6136 6137 6138 6139 6140 6141 6142 6143 6144 6145 6146 6147 6148 6149 6150 6151 6152 6153 6154 6155 6156 6157 6158 6159 6160 6161 6162 6163 6164 6165 6166 6167 6168 6169 6170 6171 6172 6173 6174 6175 6176 6177 6178 6179 6180 6181 6182 6183 6184 6185 6186 6187 6188 6189 6190 6191 6192 6193 6194 6195 6196 6197 6198 6199 6200 6201 6202 6203 6204 6205 6206 6207 6208 6209 6210 6211 6212 6213 6214 6215 6216 6217 6218 6219 6220 6221 6222 6223 6224 6225 6226 6227 6228 6229 6230 6231 6232 6233 6234 6235 6236 6237 6238 6239 6240 6241 6242 6243 6244 6245 6246 6247 6248 6249 6250 6251 6252 6253 6254 6255 6256 6257 6258 6259 6260 6261 6262 6263 6264 6265 6266 6267 6268 6269 6270 6271 6272 6273 6274 6275 6276 6277 6278 6279 6280 6281 6282 6283 6284 6285 6286 6287 6288 6289 6290 6291 6292 6293 6294 6295 6296 6297 6298 6299 6300 6301 6302 6303 6304 6305 6306 6307 6308 6309 6310 6311 6312 6313 6314 6315 6316 6317 6318 6319 6320 6321 6322 6323 6324 6325 6326 6327 6328 6329 6330 6331 6332 6333 6334 6335 6336 6337 6338 6339 6340 6341 6342 6343 6344 6345 6346 6347 6348 6349 6350 6351 6352 6353 6354 6355 6356 6357 6358 6359 6360 6361 6362 6363 6364 6365 6366 6367 6368 6369 6370 6371 6372 6373 6374 6375 6376 6377 6378 6379 6380 6381 6382 6383 6384 6385 6386 6387 6388 6389 6390 6391 6392 6393 6394 6395 6396 6397 6398 6399 6400 6401 6402 6403 6404 6405 6406 6407 6408 6409 6410 6411 6412 6413 6414 6415 6416 6417 6418 6419 6420 6421 6422 6423 6424 6425 6426 6427 6428 6429 6430 6431 6432 6433 6434 6435 6436 6437 6438 6439 6440 6441 6442 6443 6444 6445 6446 6447 6448 6449 6450 6451 6452 6453 6454 6455 6456 6457 6458 6459 6460 6461 6462 6463 6464 6465 6466 6467 6468 6469 6470 6471 6472 6473 6474 6475 6476 6477 6478 6479 6480 6481 6482 6483 6484 6485 6486 6487 6488 6489 6490 6491 6492 6493 6494 6495 6496 6497 6498 6499 6500 6501 6502 6503 6504 6505 6506 6507 6508 6509 6510 6511 6512 6513 6514 6515 6516 6517 6518 6519 6520 6521 6522 6523 6524 6525 6526 6527 6528 6529 6530 6531 6532 6533 6534 6535 6536 6537 6538 6539 6540 6541 6542 6543 6544 6545 6546 6547 6548 6549 6550 6551 6552 6553 6554 6555 6556 6557 6558 6559 6560 6561 6562 6563 6564 6565 6566 6567 6568 6569 6570 6571 6572 6573 6574 6575 6576 6577 6578 6579 6580 6581 6582 6583 6584 6585 6586 6587 6588 6589 6590 6591 6592 6593 6594 6595 6596 6597 6598 6599 6600 6601 6602 6603 6604 6605 6606 6607 6608 6609 6610 6611 6612 6613 6614 6615 6616 6617 6618 6619 6620 6621 6622 6623 6624 6625 6626 6627 6628 6629 6630 6631 6632 6633 6634 6635 6636 6637 6638 6639 6640 6641 6642 6643 6644 6645 6646 6647 6648 6649 6650 6651 6652 6653 6654 6655 6656 6657 6658 6659 6660 6661 6662 6663 6664 6665 6666 6667 6668 6669 6670 6671 6672 6673 6674 6675 6676 6677 6678 6679 6680 6681 6682 6683 6684 6685 6686 6687 6688 6689 6690 6691 6692 6693 6694 6695 6696 6697 6698 6699 6700 6701 6702 6703 6704 6705 6706 6707 6708 6709 6710 6711 6712 6713 6714 6715 6716 6717 6718 6719 6720 6721 6722 6723 6724 6725 6726 6727 6728 6729 6730 6731 6732 6733 6734 6735 6736 6737 6738 6739 6740 6741 6742 6743 6744 6745 6746 6747 6748 6749 6750 6751 6752 6753 6754 6755 6756 6757 6758 6759 6760 6761 6762 6763 6764 6765 6766 6767 6768 6769 6770 6771 6772 6773 6774 6775 6776 6777 6778 6779 6780 6781 6782 6783 6784 6785 6786 6787 6788 6789 6790 6791 6792 6793 6794 6795 6796 6797 6798 6799 6800 6801 6802 6803 6804 6805 6806 6807 6808 6809 6810 6811 6812 6813 6814 6815 6816 6817 6818 6819 6820 6821 6822 6823 6824 6825 6826 6827 6828 6829 6830 6831 6832 6833 6834 6835 6836 6837 6838 6839 6840 6841 6842 6843 6844 6845 6846 6847 6848 6849 6850 6851 6852 6853 6854 6855 6856 6857 6858 6859 6860 6861 6862 6863 6864 6865 6866 6867 6868 6869 6870 6871 6872 6873 6874 6875 6876 6877 6878 6879 6880 6881 6882 6883 6884 6885 6886 6887 6888 6889 6890 6891 6892 6893 6894 6895 6896 6897 6898 6899 6900 6901 6902 6903 6904 6905 6906 6907 6908 6909 6910 6911 6912 6913 6914 6915 6916 6917 6918 6919 6920 6921 6922 6923 6924 6925 6926 6927 6928 6929 6930 6931 6932 6933 6934 6935 6936 6937 6938 6939 6940 6941 6942 6943 6944 6945 6946 6947 6948 6949 6950 6951 6952 6953 6954 6955 6956 6957 6958 6959 6960 6961 6962 6963 6964 6965 6966 6967 6968 6969 6970 6971 6972 6973 6974 6975 6976 6977 6978 6979 6980 6981 6982 6983 6984 6985 6986 6987 6988 6989 6990 6991 6992 6993 6994 6995 6996 6997 6998 6999 7000 7001 7002 7003 7004 7005 7006 7007 7008 7009 7010 7011 7012 7013 7014 7015 7016 7017 7018 7019 7020 7021 7022 7023 7024 7025 7026 7027 7028 7029 7030 7031 7032 7033 7034 7035 7036 7037 7038 7039 7040 7041 7042 7043 7044 7045 7046 7047 7048 7049 7050 7051 7052 7053 7054 7055 7056 7057 7058 7059 7060 7061 7062 7063 7064 7065 7066 7067 7068 7069 7070 7071 7072 7073 7074 7075 7076 7077 7078 7079 7080 7081 7082 7083 7084 7085 7086 7087 7088 7089 7090 7091 7092 7093 7094 7095 7096 7097 7098 7099 7100 7101 7102 7103 7104 7105 7106 7107 7108 7109 7110 7111 7112 7113 7114 7115 7116 7117 7118 7119 7120 7121 7122 7123 7124 7125 7126 7127 7128 7129 7130 7131 7132 7133 7134 7135 7136 7137 7138 7139 7140 7141 7142 7143 7144 7145 7146 7147 7148 7149 7150 7151 7152 7153 7154 7155 7156 7157 7158 7159 7160 7161 7162 7163 7164 7165 7166 7167 7168 7169 7170 7171 7172 7173 7174 7175 7176 7177 7178 7179 7180 7181 7182 7183 7184 7185 7186 7187 7188 7189 7190 7191 7192 7193 7194 7195 7196 7197 7198 7199 7200 7201 7202 7203 7204 7205 7206 7207 7208 7209 7210 7211 7212 7213 7214 7215 7216 7217 7218 7219 7220 7221 7222 7223 7224 7225 7226 7227 7228 7229 7230 7231 7232 7233 7234 7235 7236 7237 7238 7239 7240 7241 7242 7243 7244 7245 7246 7247 7248 7249 7250 7251 7252 7253 7254 7255 7256 7257 7258 7259 7260 7261 7262 7263 7264 7265 7266 7267 7268 7269 7270 7271 7272 7273 7274 7275 7276 7277 7278 7279 7280 7281 7282 7283 7284 7285 7286 7287 7288 7289 7290 7291 7292 7293 7294 7295 7296 7297 7298 7299 7300 7301 7302 7303 7304 7305 7306 7307 7308 7309 7310 7311 7312 7313 7314 7315 7316 7317 7318 7319 7320 7321 7322 7323 7324 7325 7326 7327 7328 7329 7330 7331 7332 7333 7334 7335 7336 7337 7338 7339 7340 7341 7342 7343 7344 7345 7346 7347 7348 7349 7350 7351 7352 7353 7354 7355 7356 7357 7358 7359 7360 7361 7362 7363 7364 7365 7366 7367 7368 7369 7370 7371 7372 7373 7374 7375 7376 7377 7378 7379 7380 7381 7382 7383 7384 7385 7386 7387 7388 7389 7390 7391 7392 7393 7394 7395 7396 7397 7398 7399 7400 7401 7402 7403 7404 7405 7406 7407 7408 7409 7410 7411 7412 7413 7414 7415 7416 7417 7418 7419 7420 7421 7422 7423 7424 7425 7426 7427 7428 7429 7430 7431 7432 7433 7434 7435 7436 7437 7438 7439 7440 7441 7442 7443 7444 7445 7446 7447 7448 7449 7450 7451 7452 7453 7454 7455 7456 7457 7458 7459 7460 7461 7462 7463 7464 7465 7466 7467 7468 7469 7470 7471 7472 7473 7474 7475 7476 7477 7478 7479 7480 7481 7482 7483 7484 7485 7486 7487 7488 7489 7490 7491 7492 7493 7494 7495 7496 7497 7498 7499 7500 7501 7502 7503 7504 7505 7506 7507 7508 7509 7510 7511 7512 7513 7514 7515 7516 7517 7518 7519 7520 7521 7522 7523 7524 7525 7526 7527 7528 7529 7530 7531 7532 7533 7534 7535 7536 7537 7538 7539 7540 7541 7542 7543 7544 7545 7546 7547 7548 7549 7550 7551 7552 7553 7554 7555 7556 7557 7558 7559 7560 7561 7562 7563 7564 7565 7566 7567 7568 7569 7570 7571 7572 7573 7574 7575 7576 7577 7578 7579 7580 7581 7582 7583 7584 7585 7586 7587 7588 7589 7590 7591 7592 7593 7594 7595 7596 7597 7598 7599 7600 7601 7602 7603 7604 7605 7606 7607 7608 7609 7610 7611 7612 7613 7614 7615 7616 7617 7618 7619 7620 7621 7622 7623 7624 7625 7626 7627 7628 7629 7630 7631 7632 7633 7634 7635 7636 7637 7638 7639 7640 7641 7642 7643 7644 7645 7646 7647 7648 7649 7650 7651 7652 7653 7654 7655 7656 7657 7658 7659 7660 7661 7662 7663 7664 7665 7666 7667 7668 7669 7670 7671 7672 7673 7674 7675 7676 7677 7678 7679 7680 7681 7682 7683 7684 7685 7686 7687 7688 7689 7690 7691 7692 7693 7694 7695 7696 7697 7698 7699 7700 7701 7702 7703 7704 7705 7706 7707 7708 7709 7710 7711 7712 7713 7714 7715 7716 7717 7718 7719 7720 7721 7722 7723 7724 7725 7726 7727 7728 7729 7730 7731 7732 7733 7734 7735 7736 7737 7738 7739 7740 7741 7742 7743 7744 7745 7746 7747 7748 7749 7750 7751 7752 7753 7754 7755 7756 7757 7758 7759 7760 7761 7762 7763 7764 7765 7766 7767 7768 7769 7770 7771 7772 7773 7774 7775 7776 7777 7778 7779 7780 7781 7782 7783 7784 7785 7786 7787 7788 7789 7790 7791 7792 7793 7794 7795 7796 7797 7798 7799 7800 7801 7802 7803 7804 7805 7806 7807 7808 7809 7810 7811 7812 7813 7814 7815 7816 7817 7818 7819 7820 7821 7822 7823 7824 7825 7826 7827 7828 7829 7830 7831 7832 7833 7834 7835 7836 7837 7838 7839 7840 7841 7842 7843 7844 7845 7846 7847 7848 7849 7850 7851 7852 7853 7854 7855 7856 7857 7858 7859 7860 7861 7862 7863 7864 7865 7866 7867 7868 7869 7870 7871 7872 7873 7874 7875 7876 7877 7878 7879 7880 7881 7882 7883 7884 7885 7886 7887 7888 7889 7890 7891 7892 7893 7894 7895 7896 7897 7898 7899 7900 7901 7902 7903 7904 7905 7906 7907 7908 7909 7910 7911 7912 7913 7914 7915 7916 7917 7918 7919 7920 7921 7922 7923 7924 7925 7926 7927 7928 7929 7930 7931 7932 7933 7934 7935 7936 7937 7938 7939 
<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
   "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html lang="en" xmlns="http://www.w3.org/1999/xhtml">
  <head>
    <title>Portable Network Graphics (PNG) Specification (Second Edition)</title>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> 
    <link rel="stylesheet" href="./isostyle.css" type="text/css" />
    <style type="text/css">
    /* remove annoying green color from definition terms */
    dt {color: black}
    </style>
    <link rel="stylesheet" type="text/css" media="screen"
    href="http://www.w3.org/StyleSheets/TR/W3C-REC" />
  </head>
  <body>
    <div class="head">
      <p><a href="http://www.w3.org/"><img height="48" width="72"
      alt="W3C" src="http://www.w3.org/Icons/w3c_home" /></a></p>
      <h1 id="pagetitle">Portable Network Graphics (PNG) Specification (Second Edition)</h1>
      <h1>Information technology — Computer graphics and image processing — Portable Network Graphics (PNG): Functional specification. ISO/IEC 15948:2003 (E)</h1>
      <!--h2 id="pagesubtitle">W3C Recommendation 1 October 1996, revised 14 October 2003</h2-->
      <h2 id="pagesubtitle">W3C Recommendation 10 November 2003</h2>
      <dl>
        <dt>This version:</dt>
        <dd><a
        href="http://www.w3.org/TR/2003/REC-PNG-20031110">http://www.w3.org/TR/2003/REC-PNG-20031110</a></dd>
        <dt>Latest version:</dt>
        <dd><a
        href="http://www.w3.org/TR/PNG">http://www.w3.org/TR/PNG</a></dd>
        <dt>Previous version:</dt>
        <dd><a
        href="http://www.w3.org/TR/2003/PR-PNG-20030520">http://www.w3.org/TR/2003/PR-PNG-20030520</a></dd>
        <dt>Editor:</dt>
        <dd>David Duce, Oxford Brookes University (Second Edition)</dd>
        <dt>Authors:</dt>
        <dd>See <a href="#F-Relationship">author list</a></dd>
      </dl>
      <p>Please refer to the <a href="http://www.w3.org/2003/11/REC-PNG-20031110-errata"><strong>errata</strong></a> for this document, which may include some normative corrections.</p>

      <!--p>This document is also available in these non-normative
      packages: <a href="REC-SVG11-20030114.zip">zip archive of
      HTML</a> (without external dependencies) and <a
      href="REC-SVG11-20030114.pdf">PDF</a>.</p-->
  
      <p>See also the <a href="http://www.w3.org/Consortium/Translation/">translations</a> of this document.</p>

<p class="copyright"><a href="http://www.w3.org/Consortium/Legal/ipr-notice#Copyright"> Copyright</a> &#xa9; 2003 <a href="http://www.w3.org/"><acronym title="World Wide Web Consortium">W3C</acronym></a><sup>&#xae;</sup> (<a href="http://www.lcs.mit.edu/"><acronym title="Massachusetts Institute of Technology">MIT</acronym></a>, <a href="http://www.ercim.org/"><acronym title="European Research Consortium for Informatics and Mathematics">ERCIM</acronym></a>, <a href="http://www.keio.ac.jp/">Keio</a>), All Rights Reserved. W3C <a href="http://www.w3.org/Consortium/Legal/ipr-notice#Legal_Disclaimer">liability</a>, <a href="http://www.w3.org/Consortium/Legal/ipr-notice#W3C_Trademarks">trademark</a>, <a href="http://www.w3.org/Consortium/Legal/copyright-documents">document use</a> and <a href="http://www.w3.org/Consortium/Legal/copyright-software">software licensing</a> rules apply.</p>
    </div>
    <hr title="Separator from Header" />

    <h2 id="specabstract"><a id="abstract" name="abstract">Abstract</a></h2>
    <p>This document describes PNG (Portable Network Graphics), an extensible file format for the lossless, portable, well-compressed storage of raster images. PNG provides a patent-free replacement for GIF and can also replace many common uses of TIFF. Indexed-color, grayscale, and truecolor images are supported, plus an optional alpha channel. Sample depths range from 1 to 16 bits.</p>
 <p>PNG is designed to work well in online viewing applications, such as the World Wide Web, so it is fully streamable with a progressive display option. PNG is robust, providing both full file integrity checking and simple detection of common transmission errors. Also, PNG can store gamma and chromaticity data for improved color matching on heterogeneous platforms.</p>

 <p>This specification defines an Internet Media Type image/png.</p>

    <h2 id="status">Status of this document</h2>
<p><em>This section describes the status of this document at the time of its publication. Other documents may supersede this document. A list of current W3C publications and the latest revision of this technical report can be found in the <a href="http://www.w3.org/TR/">W3C technical reports index</a> at http://www.w3.org/TR/.</em></p> 
    
    <p>This document is the 14 October 2003 W3C
    Recommendation of the PNG specification, second edition. It is also International Standard, ISO/IEC 15948:2003. The two documents have exactly identical content except for cover page and boilerplate differences as appropriate to the two organisations.</p>

<p>This International Standard is strongly based on the W3C Recommendation 'PNG Specification Version 1.0' which was reviewed by W3C members, approved as a W3C Recommendation and published in October 1996. This second edition incorporates all known errata and clarifications. </p>

<p>A complete review of the document has been done by ISO/IEC/JTC 1/SC 24 in collaboration with W3C and the PNG development group (the original authors of the PNG 1.0 Recommendation) in order to transform that Recommendation into an ISO/IEC international standard. A major design goal during this review was to avoid changes that will invalidate existing files, editors, or viewers that conform to W3C Recommendation PNG Specification Version 1.0.</p>

<p>The PNG specification enjoys a good level of <a href="http://www.libpng.org/pub/png/pngstatus.html">implementation</a>  with good interoperability. At the time of this publication more than 180 <a href="http://www.libpng.org/pub/png/pngapvw.html">image viewers</a> could display PNG images and over 100 <a href="http://www.libpng.org/pub/png/pngaped.html">image editors</a> could read and write valid PNG files. Full support of PNG is  required  for conforming <a href="/Graphics/SVG">SVG</a> viewers; at the time of publication all eighteen <a href="/Graphics/SVG/SVG-Implementations.htm8#viewer">SVG viewers</a> had PNG support. HTML has no required image formats, but over 60 <a href="http://www.libpng.org/pub/png/pngapbr.html">HTML browsers</a> had at least basic support of PNG images.</p>

    <p>Public comments on this W3C Recommendation are welcome. 
    Please send them to the <a href="http://lists.w3.org/Archives/Public/png-group">archived</a> list <a href="mailto:png-group@w3.org">png-group@w3.org</a> .</p>

    <p>The latest information regarding <a rel="disclosure"
    href="http://www.w3.org/Graphics/PNG/Disclosures">patent
    disclosures</a> related to this document is available on the
    Web. As of this publication, the PNG Group are not
    aware of any royalty-bearing patents they believe to be
    essential to PNG.</p>
    
    <p>This document has been produced by ISO/IEC JTC1 SC24 and the PNG Group as part of the <a
    href="http://www.w3.org/Graphics/Activity">Graphics
    Activity</a> within the <a href="http://www.w3.org/Interaction/">W3C
    Interaction Domain</a>. </p>
    
    <!-- removed p>A list of current W3C Recommendations and
    other technical documents can be found at <a
	href="http://www.w3.org/TR/">http://www.w3.org/TR/</a>.
	W3C  publications may be updated, replaced, or obsoleted by other 
  documents at any time.
    </p-->
    
    <div><p><strong>Note:</strong> To provide the highest quality images, this specification uses SVG diagrams with a PNG fallback using the HTML object element. SVG-enabled browsers will see the SVG figures with selectable text, other browsers will display the raster PNG version.</p>
<p>W3C is aware that there is a <a href="http://bugzilla.mozilla.org/show_bug.cgi?id=133567">known incompatibility</a> between the unsupported beta of Adobe SVG plugin for Linux and Mozilla versions greater than 0.9.9 due to changes in the plug-in API, causing a browser crash. Therefore, a normative <a href="./index-noobject.html">PNG-only alternative version</a> is available that does not use an object element. The two versions are otherwise identical.</p></div>

    <h3 id="AvailableLanguages">Available languages</h3>
    <p>The English version of this specification is the only
    normative version. However, for translations in other languages
    see <a
    href="http://www.w3.org/Consortium/Translation/">
    http://www.w3.org/Consortium/Translation/</a>.</p>

    <div class="toc">
      <h2><a id="minitoc" name="minitoc">Table of Contents</a></h2>
      <ul class="toc">
        <!--li class="tocline1"-->
<li class='Contents'><a class='Href' href='#1Scope'>1
Scope</a></li>

<li class='Contents'><a class='Href' href='#2NormRefs'>2
Normative references</a></li>

<li class='Contents'><a class='Href' href='#3Defsandabbrevs'>3
Terms, definitions, and abbreviated terms</a> 

<ul>
<li class='Contents'><a class='Href' href='#3Definitions'>3.1
Definitions</a></li>

<li class='Contents'><a class='Href' href='#3Abbreviations'>3.2
Abbreviated terms</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#4Concepts'>4
Concepts</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#4Concepts.Sourceimage'>4.1 Images</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.ColourSpaces'>4.2 Colour spaces</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.PNGImageTransformation'>4.3 Reference image to PNG
image transformation</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#4Concepts.Introduction'>4.3.1 Introduction</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.Implied-alpha'>4.3.2 Alpha separation</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.Indexing'>4.3.3 Indexing</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.RGBMerging'>4.3.4 RGB merging</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.Alpha-indexing'>4.3.5 Alpha compaction</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.Scaling'>4.3.6 Sample depth scaling</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href=
'#4Concepts.PNGImage'>4.4 PNG image</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.Encoding'>4.5 Encoding the PNG image</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#4Concepts.EncodingIntro'>4.5.1 Introduction</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.EncodingPassAbs'>4.5.2 Pass extraction</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.EncodingScanlineAbs'>4.5.3 Scanline
serialization</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.EncodingFiltering'>4.5.4 Filtering</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.EncodingCompression'>4.5.5 Compression</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.EncodingChunking'>4.5.6 Chunking</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href=
'#4Concepts.AncillInfo'>4.6 Additional information</a></li>

<li class='Contents'><a class='Href' href='#4Concepts.Format'>4.7
PNG datastream</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#4Concepts.FormatChunks'>4.7.1 Chunks</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.FormatTypes'>4.7.2 Chunk types</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#4Concepts.Errors'>4.8
Error handling</a></li>

<li class='Contents'><a class='Href' href=
'#4Concepts.Registration'>4.9 Extension and registration</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#5DataRep'>5
Datastream structure</a> 

<ul>
<li class='Contents'><a class='Href' href='#5Introduction'>5.1
Introduction</a></li>

<li class='Contents'><a class='Href' href=
'#5PNG-file-signature'>5.2 PNG signature</a></li>

<li class='Contents'><a class='Href' href='#5Chunk-layout'>5.3
Chunk layout</a></li>

<li class='Contents'><a class='Href' href=
'#5Chunk-naming-conventions'>5.4 Chunk naming
conventions</a></li>

<li class='Contents'><a class='Href' href='#5CRC-algorithm'>5.5
Cyclic Redundancy Code algorithm</a></li>

<li class='Contents'><a class='Href' href='#5ChunkOrdering'>5.6
Chunk ordering</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#6Transformation'>6
Reference image to PNG image transformation</a> 

<ul>
<li class='Contents'><a class='Href' href='#6Colour-values'>6.1
Colour types and values</a></li>

<li class='Contents'><a class='Href' href=
'#6AlphaRepresentation'>6.2 Alpha representation</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#7Transformation'>7
Encoding the PNG image as a PNG datastream</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#7Integers-and-byte-order'>7.1 Integers and byte order</a></li>

<li class='Contents'><a class='Href' href='#7Scanline'>7.2
Scanlines</a></li>

<li class='Contents'><a class='Href' href='#7Filtering'>7.3
Filtering</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#8Interlace'>8
Interlacing and pass extraction</a> 

<ul>
<li class='Contents'><a class='Href' href='#8InterlaceIntro'>8.1
Introduction</a></li>

<li class='Contents'><a class='Href' href=
'#8InterlaceMethods'>8.2 Interlace methods</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#9Filters'>9
Filtering</a> 

<ul>
<li class='Contents'><a class='Href' href='#9FtIntro'>9.1 Filter
methods and filter types</a></li>

<li class='Contents'><a class='Href' href='#9Filter-types'>9.2
Filter types for filter method 0</a></li>

<li class='Contents'><a class='Href' href=
'#9Filter-type-3-Average'>9.3 Filter type 3: Average</a></li>

<li class='Contents'><a class='Href' href=
'#9Filter-type-4-Paeth'>9.4 Filter type 4: Paeth</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#10Compression'>10
Compression</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#10CompressionCM0'>10.1 Compression method 0</a></li>

<li class='Contents'><a class='Href' href=
'#10CompressionFSL'>10.2 Compression of the sequence of filtered
scanlines</a></li>

<li class='Contents'><a class='Href' href=
'#10CompressionOtherUses'>10.3 Other uses of compression</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#11Chunks'>11 Chunk
specifications</a> 

<ul>
<li class='Contents'><a class='Href' href='#11Introduction'>11.1
Introduction</a></li>

<li class='Contents'><a class='Href' href=
'#11Critical-chunks'>11.2 Critical chunks</a> 

<ul>
<li class='Contents'><a class='Href' href='#11CcGen'>11.2.1
General</a></li>

<li class='Contents'><a class='Href' href='#11IHDR'>11.2.2 <span
class='chunk'>IHDR</span> Image header</a></li>

<li class='Contents'><a class='Href' href='#11PLTE'>11.2.3 <span
class='chunk'>PLTE</span> Palette</a></li>

<li class='Contents'><a class='Href' href='#11IDAT'>11.2.4 <span
class='chunk'>IDAT</span> Image data</a></li>

<li class='Contents'><a class='Href' href='#11IEND'>11.2.5 <span
class='chunk'>IEND</span> Image trailer</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href=
'#11Ancillary-chunks'>11.3 Ancillary chunks</a> 

<ul>
<li class='Contents'><a class='Href' href='#11AcGen'>11.3.1
General</a></li>

<li class='Contents'><a class='Href' href='#11transinfo'>11.3.2
Transparency information</a> 

<ul>
<li class='Contents'><a class='Href' href='#11tRNS'>11.3.2.1
<span class='chunk'>tRNS</span> Transparency</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href=
'#11addnlcolinfo'>11.3.3 Colour space information</a> 

<ul>
<li class='Contents'><a class='Href' href='#11cHRM'>11.3.3.1
<span class='chunk'>cHRM</span> Primary chromaticities and white
point</a></li>

<li class='Contents'><a class='Href' href='#11gAMA'>11.3.3.2
<span class='chunk'>gAMA</span> Image gamma</a></li>

<li class='Contents'><a class='Href' href='#11iCCP'>11.3.3.3
<span class='chunk'>iCCP</span> Embedded ICC profile</a></li>

<li class='Contents'><a class='Href' href='#11sBIT'>11.3.3.4
<span class='chunk'>sBIT</span> Significant bits</a></li>

<li class='Contents'><a class='Href' href='#11sRGB'>11.3.3.5
<span class='chunk'>sRGB</span> Standard RGB colour
space</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#11textinfo'>11.3.4
Textual information</a> 

<ul>
<li class='Contents'><a class='Href' href='#11textIntro'>11.3.4.1
Introduction</a></li>

<li class='Contents'><a class='Href' href='#11keywords'>11.3.4.2
Keywords and text strings</a></li>

<li class='Contents'><a class='Href' href='#11tEXt'>11.3.4.3
<span class='chunk'>tEXt</span> Textual data</a></li>

<li class='Contents'><a class='Href' href='#11zTXt'>11.3.4.4
<span class='chunk'>zTXt</span> Compressed textual data</a></li>

<li class='Contents'><a class='Href' href='#11iTXt'>11.3.4.5
<span class='chunk'>iTXt</span> International textual
data</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#11addnlsiinfo'>11.3.5
Miscellaneous information</a> 

<ul>
<li class='Contents'><a class='Href' href='#11bKGD'>11.3.5.1
<span class='chunk'>bKGD</span> Background colour</a></li>

<li class='Contents'><a class='Href' href='#11hIST'>11.3.5.2
<span class='chunk'>hIST</span> Image histogram</a></li>

<li class='Contents'><a class='Href' href='#11pHYs'>11.3.5.3
<span class='chunk'>pHYs</span> Physical pixel
dimensions</a></li>

<li class='Contents'><a class='Href' href='#11sPLT'>11.3.5.4
<span class='chunk'>sPLT</span> Suggested palette</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href=
'#11timestampinfo'>11.3.6 Time stamp information</a> 

<ul>
<li class='Contents'><a class='Href' href='#11tIME'>11.3.6.1
<span class='chunk'>tIME</span> Image last-modification
time</a></li>

</ul>
</li>
</ul>
</li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#12Encoders'>12 PNG
Encoders</a> 

<ul>
<li class='Contents'><a class='Href' href='#12Introduction'>12.1
Introduction</a></li>

<li class='Contents'><a class='Href' href=
'#12Encoder-gamma-handling'>12.2 Encoder gamma handling</a></li>

<li class='Contents'><a class='Href' href=
'#12Encoder-colour-handling'>12.3 Encoder colour
handling</a></li>

<li class='Contents'><a class='Href' href=
'#12Alpha-channel-creation'>12.4 Alpha channel creation</a></li>

<li class='Contents'><a class='Href' href=
'#12Sample-depth-scaling'>12.5 Sample depth scaling</a></li>

<li class='Contents'><a class='Href' href=
'#12Suggested-palettes'>12.6 Suggested palettes</a></li>

<li class='Contents'><a class='Href' href='#12Interlacing'>12.7
Interlacing</a></li>

<li class='Contents'><a class='Href' href=
'#12Filter-selection'>12.8 Filter selection</a></li>

<li class='Contents'><a class='Href' href='#12Compression'>12.9
Compression</a></li>

<li class='Contents'><a class='Href' href=
'#12Text-chunk-processing'>12.10 Text chunk processing</a></li>

<li class='Contents'><a class='Href' href=
'#12Chunk-processing'>12.11 Chunking</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#12Use-of-private-chunks'>12.11.1 Use of private chunks</a></li>

<li class='Contents'><a class='Href' href=
'#12Private-type-and-method-codes'>12.11.2 Private type and
method codes</a></li>

<li class='Contents'><a class='Href' href='#12Ancillary'>12.11.3
Ancillary chunks</a></li>
</ul>
</li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#13Decoders'>13 PNG
decoders and viewers</a> 

<ul>
<li class='Contents'><a class='Href' href='#13Introduction'>13.1
Introduction</a></li>

<li class='Contents'><a class='Href' href=
'#13Decoders.Errors'>13.2 Error handling</a></li>

<li class='Contents'><a class='Href' href=
'#13Error-checking'>13.3 Error checking</a></li>

<li class='Contents'><a class='Href' href=
'#13Security-considerations'>13.4 Security
considerations</a></li>

<li class='Contents'><a class='Href' href='#13Chunking'>13.5
Chunking</a></li>

<li class='Contents'><a class='Href' href=
'#13Pixel-dimensions'>13.6 Pixel dimensions</a></li>

<li class='Contents'><a class='Href' href=
'#13Text-chunk-processing'>13.7 Text chunk processing</a></li>

<li class='Contents'><a class='Href' href='#13Decompression'>13.8
Decompression</a></li>

<li class='Contents'><a class='Href' href='#13Filtering'>13.9
Filtering</a></li>

<li class='Contents'><a class='Href' href=
'#13Progressive-display'>13.10 Interlacing and progressive
display</a></li>

<li class='Contents'><a class='Href' href=
'#13Truecolour-image-handling'>13.11 Truecolour image
handling</a></li>

<li class='Contents'><a class='Href' href=
'#13Sample-depth-rescaling'>13.12 Sample depth rescaling</a></li>

<li class='Contents'><a class='Href' href=
'#13Decoder-gamma-handling'>13.13 Decoder gamma handling</a></li>

<li class='Contents'><a class='Href' href=
'#13Decoder-colour-handling'>13.14 Decoder colour
handling</a></li>

<li class='Contents'><a class='Href' href=
'#13Background-colour'>13.15 Background colour</a></li>

<li class='Contents'><a class='Href' href=
'#13Alpha-channel-processing'>13.16 Alpha channel
processing</a></li>

<li class='Contents'><a class='Href' href=
'#13Histogram-and-suggested-palette-usage'>13.17
Histogram and suggested palette usage</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#14EditorsExt'>14
Editors and extensions</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#14Additional-chunk-types'>14.1 Additional chunk types</a></li>

<li class='Contents'><a class='Href' href='#14Ordering'>14.2
Behaviour of PNG editors</a></li>

<li class='Contents'><a class='Href' href=
'#14Ordering-of-chunks'>14.3 Ordering of chunks</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#14Ordering-of-critical-chunks'>14.3.1 Ordering of critical
chunks</a></li>

<li class='Contents'><a class='Href' href=
'#14Ordering-of-ancillary-chunks'>14.3.2 Ordering of ancillary
chunks</a></li>
</ul>
</li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#15Conformance'>15
Conformance</a> 

<ul>
<li class='Contents'><a class='Href' href='#15ConfIntro'>15.1
Introduction</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#15ConfObjectives'>15.1.1 Objectives</a></li>

<li class='Contents'><a class='Href' href='#15ConfScope'>15.1.2
Scope</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href=
'#15ConformanceConf'>15.2 Conformance conditions</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#15FileConformance'>15.2.1 Conformance of PNG
datastreams</a></li>

<li class='Contents'><a class='Href' href=
'#15ConformanceEncoder'>15.2.2 Conformance of PNG
encoders</a></li>

<li class='Contents'><a class='Href' href=
'#15ConformanceDecoder'>15.2.3 Conformance of PNG
decoders</a></li>

<li class='Contents'><a class='Href' href=
'#15ConformanceEditor'>15.2.4 Conformance of PNG editors</a></li>
</ul>
</li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#A-Conventions'>Annex
A File conventions and Internet media type</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#A-File-name-extension'>A.1 File name extension</a></li>

<li class='Contents'><a class='Href' href='#A-Media-type'>A.2
Internet media type</a></li>

<li class='Contents'><a class='Href' href=
'#A-Macintosh-file-layout'>A.3 Macintosh file layout</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href=
'#B-NewChunksAppendix'>Annex B Guidelines for new chunk
types</a></li>

<li class='Contents'><a class='Href' href=
'#C-GammaAppendix'>Annex C Gamma and chromaticity</a></li>

<li class='Contents'><a class='Href' href='#D-CRCAppendix'>Annex
D Sample Cyclic Redundancy Code implementation</a></li>

<li class='Contents'><a class='Href' href='#E-Resources'>Annex E
Online resources</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#E-Intro'>Introduction</a></li>

<li class='Contents'><a class='Href' href=
'#E-Archive-sites'>Archive sites</a></li>

<li class='Contents'><a class='Href' href=
'#E-icc-profile-specs'>ICC profile specifications</a></li>

<li class='Contents'><a class='Href' href='#E-PNG-home-page'>PNG
web site</a></li>

<li class='Contents'><a class='Href' href=
'#E-Sample-implementation'>Sample implementation and test
images</a></li>

<li class='Contents'><a class='Href' href='#E-Email'>Electronic
mail</a></li>
</ul>
</li>

<li class='Contents'><a class='Href' href='#F-Relationship'>Annex
F Relationship to W3C PNG</a> 

<ul>
<li class='Contents'><a class='Href' href='#F-Editor10'>Editor
(Version 1.0)</a></li>

<li class='Contents'><a class='Href' href='#F-Editor12'>Editor
(Versions 1.1 and 1.2)</a></li>

<li class='Contents'><a class='Href' href=
'#F-ContribEditor10'>Contributing Editor (Version 1.0)</a></li>

<li class='Contents'><a class='Href' href=
'#F-ContribEditor12'>Contributing Editor (Versions 1.1 and 1.2)</a></li>

<li class='Contents'><a class='Href' href='#F-Authors'>Authors
(Versions 1.0, 1.1, and 1.2 combined)</a></li>

<li class='Contents'><a class='Href' href='#F-ChangeList'>List of
changes between W3C Recommendation PNG Specification Version 1.0
and this International Standard</a> 

<ul>
<li class='Contents'><a class='Href' href=
'#F-EditorialChanges'>Editorial changes</a></li>

<li class='Contents'><a class='Href' href=
'#F-TechnicalChanges'>Technical changes</a></li>
</ul>
</li>
</ul>
</li>

<li class='Contents'><a class='Href' href=
'#G-References'>Bibliography</a></li>
</ul>
</div>

<!-- *********************************************************************

FROM HERE ON THIS FILE IS IDENTICAL TO THE ISO VERSION
with these exceptions:

- id added to any headings that did not have one, to comply with pubrules and allow indexing into the document
- URL for this document updated in Annex E and the words " [to be completed when published]" removed

**************************************************************************  -->

<h1><a name="Introduction">Introduction</a></h1>

<p></p>

<p>The design goals for this International Standard were:</p>

<ol>
<li>Portability: encoding, decoding, and transmission should be
software and hardware platform independent.</li>

<li>Completeness: it should be possible to represent truecolour,
indexed-colour, and greyscale images, in each case with the
option of transparency, colour space information, and ancillary
information such as textual comments.</li>

<li>Serial encode and decode: it should be possible for
datastreams to be generated serially and read serially, allowing
the datastream format to be used for on-the-fly generation and
display of images across a serial communication channel.</li>

<li>Progressive presentation: it should be possible to transmit
datastreams so that an approximation of the whole image can be
presented initially, and progressively enhanced as the datastream
is received.</li>

<li>Robustness to transmission errors: it should be possible to
detect datastream transmission errors reliably.</li>

<li>Losslessness: filtering and compression should preserve all
information.</li>

<li>Performance: any filtering, compression, and progressive
image presentation should be aimed at efficient decoding and
presentation. Fast encoding is a less important goal than fast
decoding. Decoding speed may be achieved at the expense of
encoding speed.</li>

<li>Compression: images should be compressed effectively,
consistent with the other design goals.</li>

<li>Simplicity: developers should be able to implement the
standard easily.</li>

<li>Interchangeability: any standard-conforming PNG decoder shall
be capable of reading all conforming PNG datastreams.</li>

<li>Flexibility: future extensions and private additions should
be allowed for without compromising the interchangeability of
standard PNG datastreams.</li>

<li>Freedom from legal restrictions: no algorithms should be used
that are not freely available.</li>
</ol>


<h1><a name="1Scope">1 Scope</a></h1>

<p>This International Standard specifies a datastream and an
associated file format, Portable Network Graphics (PNG,
pronounced "ping"), for a lossless, portable, compressed
individual computer graphics image transmitted across the
Internet. Indexed-colour, greyscale, and truecolour images are
supported, with optional transparency. Sample depths range from 1
to 16 bits. PNG is fully streamable with a progressive display
option. It is robust, providing both full file integrity checking
and simple detection of common transmission errors. PNG can store
gamma and chromaticity data as well as a full ICC colour profile
for accurate colour matching on heterogenous platforms. This
Standard defines the Internet Media type "image/png". The
datastream and associated file format have value outside of the
main design goal.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="2NormRefs">2 Normative references</a></h1>

<p>The following normative documents contain provisions which,
through reference in this text, constitute provisions of this
International Standard. For dated references, subsequent
amendments to, or revisions of, any of these publications do not
apply. However, parties to agreements based on this International
Standard are encouraged to investigate the possibility of
applying the most recent editions of the normative documents
indicated below. For undated references, the latest edition of
the normative document referred to applies. Members of ISO and
IEC maintain registers of currently valid International
Standards.</p>

<p class="NormRefDef"><a name="2-ISO-639">ISO 639:1988</a>,
<i>Code for the representation of names of languages</i>.</p>

<p class="NormRefDef"><a name="2-ISO-646">ISO/IEC 646:1991</a>,
<i>International Organization for Standardization, Information
technology &mdash; ISO 7-bit coded character set for information
interchange</i>.</p>

<p class="NormRefDef"><a name="2-ISO-3309">ISO/IEC 3309:1993</a>,
<i>Information Technology &mdash; Telecommunications and
information exchange between systems &mdash; High-level data link
control (HDLC) procedures &mdash; Frame structure</i>.</p>

<p class="NormRefDef"><a name="2-ISO-8859-1">ISO/IEC
8859-1:1998</a>, <i>Information technology &mdash; 8-bit
single-byte coded graphic character sets &mdash; Part 1: Latin
alphabet No. 1</i>.<br class="xhtml" />
 For convenience, here is a non-normative  <a href="iso_8859-1.txt">sample text file</a> 
 describing the codes and associated character names.</p>

<p class="NormRefDef"><a name="2-ISO-9899">ISO/IEC
9899:1990(R1997)</a>, <i>Programming languages &mdash; C</i>.</p>

<p class="NormRefDef"><a name="2-ISO-10646-1">ISO/IEC
10646-1:1993/AMD.2</a>, <i>Information technology &mdash;
Universal Multiple-Octet Coded Character Sets (UCS) &mdash; Part
1: Architecture and Basic Multilingual Plane</i>.</p>

<p class="NormRefDef"><a name="2-IEC-61966-2-1">IEC
61966-2-1</a>, <i>Multimedia systems and equipment &mdash; Colour
measurement and management &mdash; Part 2-1: Default RGB colour
space &mdash; sRGB,</i> available at <code><a href=
"http://www.iec.ch">http://www.iec.ch/</a></code>.</p>

<p class="NormRefDef"><a name="2-CIE-15.2">CIE-15.2</a>, CIE,
"Colorimetry, Second Edition". CIE Publication 15.2-1986. ISBN
3-900-734-00-3.</p>

<p class="NormRefDef"><a name="2-ICC-1">ICC-1</a>, International
Color Consortium, "Specification ICC.1: 1998-09, File Format for
Color Profiles", 1998, available at <code><a href=
"http://www.color.org/">http://www.color.org/</a></code></p>

<p class="NormRefDef"><a name="2-ICC-1A">ICC-1A</a>,
International Color Consortium, "Specification ICC.1A: 1999-04,
Addendum 2 to ICC.1: 1998-09", 1999, available at <code><a href=
"http://www.color.org/">http://www.color.org/</a></code></p>

<p class="NormRefDef"><a name="2-RFC-1123">RFC-1123</a>, Braden,
R., Editor, "Requirements for Internet Hosts &mdash; Application
and Support", STD 3, RFC 1123, USC/Information Sciences
Institute, October 1989.<br class="xhtml" />
 <code><a href=
"http://www.ietf.org/rfc/rfc1123.txt">http://www.ietf.org/rfc/rfc1123.txt</a></code></p>

<p class="NormRefDef"><a name="2-RFC-1950">RFC-1950</a>, Deutsch,
P. and Gailly, J-L., "ZLIB Compressed Data Format Specification
version 3.3", RFC 1950, Aladdin Enterprises, May 1996.<br class="xhtml" />
 <code><a href=
"http://www.ietf.org/rfc/rfc1950.txt">http://www.ietf.org/rfc/rfc1950.txt</a></code></p>

<p class="NormRefDef"><a name="2-RFC-1951">RFC-1951</a>, Deutsch,
P., "DEFLATE Compressed Data Format Specification version 1.3",
RFC 1951, Aladdin Enterprises, May 1996.<br class="xhtml" />
 <code><a href=
"http://www.ietf.org/rfc/rfc1951.txt">http://www.ietf.org/rfc/rfc1951.txt</a></code></p>

<p class="NormRefDef"><a name="2-RFC-2045">RFC-2045</a>, Freed,
N. and Borenstein, N. , "MIME (Multipurpose Internet Mail
Extensions) Part One: Format of Internet Message Bodies", RFC
2045, Innosoft, First Virtual, November 1996.<br class="xhtml" />
 <code><a href=
"http://www.ietf.org/rfc/rfc2045.txt">http://www.ietf.org/rfc/rfc2045.txt</a></code></p>

<p class="NormRefDef"><a name="2-RFC-2048">RFC-2048</a>, Freed,
N., Klensin, J. and Postel, J., "Multipurpose Internet Mail
Extensions (MIME) Part Four: Registration Procedures", RFC 2048,
Innosoft, MCI, ISI, November 1996.<br class="xhtml" />
 <code><a href=
"http://www.ietf.org/rfc/rfc2048.txt">http://www.ietf.org/rfc/rfc2048.txt</a></code></p>

<p class="NormRefDef"><a name="2-RFC-3066">RFC-3066</a>,
Alvestrand, H., "Tags for the Identification of Languages", RFC
3066, Cisco Systems, January 2001. (Obsoletes RFC 1766.)<br class="xhtml" />
 <code><a href=
"http://www.ietf.org/rfc/rfc3066.txt">http://www.ietf.org/rfc/rfc3066.txt</a></code></p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="3Defsandabbrevs">3 Terms, definitions, and
abbreviated terms</a></h1>

<h2><a name="3Definitions">3.1 Definitions</a></h2>

<p>For the purposes of this International Standard the following
definitions apply.</p>

<dl>
<dt><a name="3alpha">3.1.1 alpha</a></dt>

<dd>a value representing a <a href="#3pixel"><span class=
"Definition">pixel's</span></a> degree of opacity. The more
opaque a pixel, the more it hides the background against which
the image is presented. Zero alpha represents a completely
transparent pixel, maximum alpha represents a completely opaque
pixel.</dd>

<dt><a name="3alphaCompaction">3.1.2 alpha compaction</a></dt>

<dd>an implicit representation of transparent <a href=
"#3pixel"><span class="Definition">pixels</span></a>. If every
pixel with a specific colour or <a href="#3greyscale"><span
class="Definition">greyscale</span></a> value is fully
transparent and all other pixels are fully opaque, the <a href=
"#3alpha"><span class="Definition">alpha</span></a> <a href=
"#3channel"><span class="Definition">channel</span></a> may be
represented implicitly.</dd>

<dt><a name="3alphaSeparation">3.1.3 alpha separation</a></dt>

<dd>separating an <a href="#3alpha"><span class=
"Definition">alpha</span></a> <a href="#3channel"><span class=
"Definition">channel</span></a> in which every <a href=
"#3pixel"><span class="Definition">pixel</span></a> is fully
opaque; all alpha values are the maximum value.
The fact that all pixels are fully opaque is represented implicitly.
</dd>

<dt><a name="3alphaTable">3.1.4 alpha table</a></dt>

<dd>indexed table of <a href="#3alpha"><span class=
"Definition">alpha</span></a> <a href="#3sample"><span class=
"Definition">sample</span></a> values, which in an <a href=
"#3indexedColour"><span class=
"Definition">indexed-colour</span></a> image defines the alpha
sample values of the <a href="#3referenceImage"><span class=
"Definition">reference image</span></a>. The alpha table has the
same number of entries as the <a href="#3palette"><span class=
"Definition">palette</span></a>.</dd>

<dt><a name="3ancillaryChunk">3.1.5 ancillary chunk</a></dt>

<dd>class of <a href="#3chunk"><span class=
"Definition">chunk</span></a> that provides additional
information. A <a href="#3PNGdecoder"><span class=
"Definition">PNG decoder</span></a>, without processing an
ancillary chunk, can still produce a meaningful image, though not
necessarily the best possible image. 
<!-- agreed: don't need to define a bit -->
</dd>

<dt><a name="3bitDepth">3.1.6 bit depth</a></dt>

<dd>for <a href="#3indexedColour"><span class=
"Definition">indexed-colour</span></a> images, the number of bits
per <a href="#3palette"><span class=
"Definition">palette</span></a> index. For other images, the
number of bits per <a href="#3sample"><span class=
"Definition">sample</span></a> in the image. This is the value
that appears in the <a href="#11IHDR"><span class=
"chunk">IHDR</span></a> <a href="#3chunk"><span class=
"Definition">chunk</span></a>.</dd>

<dt><a name="3byte">3.1.7 byte</a></dt>

<dd>8 bits; also called an octet. The highest bit (value 128) of
a byte is numbered bit 7; the lowest bit (value 1) is numbered
bit 0.</dd>

<dt><a name="3byteOrder">3.1.8 byte order</a></dt>

<dd>ordering of <a href="#3byte"><span class=
"Definition">bytes</span></a> for multi-byte data values within a
<a href="#3PNGfile"><span class="Definition">PNG file</span></a>
or <a href="#3PNGdatastream"><span class="Definition">PNG
datastream</span></a>. PNG uses <a href=
"#3networkByteOrder"><span class="Definition">network byte
order</span></a>.</dd>

<dt><a name="3channel">3.1.9 channel</a></dt>

<dd>array of all per-<a href="#3pixel"><span class=
"Definition">pixel</span></a> information of a particular kind
within a <a href="#3referenceImage"><span class=
"Definition">reference image</span></a>. There are five kinds of
information: red, green, blue, <a href="#3greyscale"><span class=
"Definition">greyscale</span></a>, and <a href="#3alpha"><span
class="Definition">alpha</span></a>. For example the alpha
channel is the array of alpha values within a reference
image.</dd>

<dt><a name="3chromaticity">3.1.10 chromaticity (CIE)</a></dt>

<dd>pair of values <i>x,y</i> that precisely specify a colour,
except for the brightness information.</dd>

<dt><a name="3chunk">3.1.11 chunk</a></dt>

<dd>section of a <a href="#3PNGdatastream"><span class=
"Definition">PNG datastream</span></a>. Each chunk has a chunk
type. Most chunks also include data. The format and meaning of
the data within the chunk are determined by the chunk type.
Each chunk is either a 
<a href="#3criticalChunk"><span class=
"Definition">critical chunk</span></a> or an <a href=
"#3ancillaryChunk"><span class=
"Definition">ancillary chunk</span></a>.
</dd>

<dt><a name="3colourType">3.1.12 colour type</a></dt>

<dd>value denoting how colour and <a href="#3alpha"><span class=
"Definition">alpha</span></a> are specified in the <a href=
"#3PNGimage"><span class="Definition">PNG image</span></a>.
Colour types are sums of the following values: 1 (<a href=
"#3palette"><span class="Definition">palette</span></a> used), 2
(<a href="#3truecolour"><span class=
"Definition">truecolour</span></a> used), 4 (alpha used). The
permitted values of colour type are 0, 2, 3, 4, and 6.</dd>

<dt><a name="3composite">3.1.13 composite (verb)</a></dt>

<dd>to form an image by merging a foreground image and a
background image, using transparency information to determine
where and to what extent the background should be visible. The
foreground image is said to be "composited against" the
background.</dd>

<dt><a name="3criticalChunk">3.1.14 critical chunk</a></dt>

<dd><a href="#3chunk"><span class="Definition">chunk</span></a>
that <!--must be understood and processed by the decoder-->
 shall be understood and processed by the decoder in order to
produce a meaningful image from a <a href="#3PNGdatastream"><span
class="Definition">PNG datastream</span></a>.</dd>

<dt><a name="3datastream">3.1.15 datastream</a></dt>

<dd>sequence of <a href="#3byte"><span class=
"Definition">bytes</span></a>. This term is used rather than
"file" to describe a byte sequence that may be only a portion of
a file. It is also used to emphasize that the sequence of bytes
might be generated and consumed "on the fly", never appearing in
a stored file at all.</dd>

<dt><a name="3deflate">3.1.16 deflate</a></dt>

<dd>name of a particular compression algorithm. This algorithm is
used, in compression mode 0, in conforming <a href=
"#3PNGdatastream"><span class="Definition">PNG
datastreams</span></a>. Deflate is a member of the <a href=
"#3LZ77"><span class="Definition">LZ77</span></a> family of
compression methods. It is defined in <a href="#2-RFC-1951"><span
class="NormRef">[RFC-1951]</span></a>.</dd>


<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->

<dt><a name="3deliveredImage">3.1.17 delivered image</a></dt>

<dd>image constructed from a decoded <a href=
"#3PNGdatastream"><span class="Definition">PNG
datastream</span></a>.</dd>

<dt><a name="3filter">3.1.18 filter</a></dt>

<dd>transformation applied to an array of <a href=
"#3scanline"><span class="Definition">scanlines</span></a> with
the aim of improving their compressibility. PNG uses only
lossless (reversible) filter algorithms.</dd>

<dt><a name="3frameBuffer">3.1.19 frame buffer</a></dt>

<dd>the final digital storage area for the image shown by most
types of computer display. Software causes an image to appear on
screen by loading the image into the frame buffer.</dd>

<dt><a name="3gamma">3.1.20 gamma</a></dt>

<dd>exponent that describes approximations to certain non-linear
transfer functions encountered in image capture and reproduction.
Within this International Standard, gamma is the exponent in the
transfer function from <tt>display_output</tt> to
<tt>image_sample</tt>
<pre>
<tt>image_sample = display_output<sup>gamma</sup></tt>
</pre>
where both <tt>display_output</tt> and <tt>image_sample</tt>
are scaled to the range 0 to 1.
</dd>

<dt><a name="3greyscale">3.1.21 greyscale</a></dt>

<dd>image representation in which each <a href="#3pixel"><span
class="Definition">pixel</span></a> is defined by a single <a
href="#3sample"><span class="Definition">sample</span></a> of
colour information, representing overall <a href=
"#3luminance"><span class="Definition">luminance</span></a> (on a
scale from black to white), and optionally an <a href=
"#3alpha"><span class="Definition">alpha</span></a> sample (in
which case it is called greyscale with alpha).</dd>

<dt><a name="3imageData">3.1.22 image data</a></dt>

<dd>1-dimensional array of <a href="#3scanline"><span class=
"Definition">scanlines</span></a> within an image.</dd>

<dt><a name="3indexedColour">3.1.23 indexed-colour</a></dt>

<dd>image representation in which each <a href="#3pixel"><span
class="Definition">pixel</span></a> of the original image is
represented by a single index into a <a href="#3palette"><span
class="Definition">palette</span></a>. The selected palette entry
defines the actual colour of the pixel.</dd>

<dt><a name="3indexing">3.1.24 indexing</a></dt>

<dd>representing an image by a <a href="#3palette"><span class=
"Definition">palette</span></a>, an <a href="#3alphaTable"><span
class="Definition">alpha table</span></a>, and an array of
indices pointing to entries in the palette and alpha table.</dd>

<dt><a name="3interlacedPNGimage">3.1.25 interlaced PNG
image</a></dt>

<dd>sequence of <a href="#3reducedImage"><span class=
"Definition">reduced images</span></a> generated from the <a
href="#3PNGimage"><span class="Definition">PNG image</span></a>
by <a href="#3passExtraction"><span class="Definition">pass
extraction</span></a>.</dd>

<dt><a name="3losslessCompression">3.1.26 lossless
compression</a></dt>

<dd>method of data compression that permits reconstruction of the
original data exactly, bit-for-bit.</dd>

<dt><a name="3lossyCompression">3.1.27 lossy compression</a></dt>

<dd>method of data compression that permits reconstruction of the
original data approximately, rather than exactly.</dd>

<dt><a name="3luminance">3.1.28 luminance</a></dt>

<dd>formal definition of luminance is in <a href=
"#2-CIE-15.2"><span class="NormRef">[CIE-15.2]</span></a>.
Informally it is the perceived brightness, or <a href=
"#3greyscale"><span class="Definition">greyscale</span></a>
level, of a colour. Luminance and <a href="#3chromaticity"><span
class="Definition">chromaticity</span></a> together fully define
a perceived colour.</dd>

<dt><a name="3LZ77">3.1.29 LZ77</a></dt>

<dd>data compression algorithm described by Ziv and Lempel in
their 1977 paper <a href="#G-ZL"><span class=
"bibref">[ZL]</span></a>.</dd>

<dt><a name="3networkByteOrder">3.1.30 network byte
order</a></dt>

<dd><a href="#3byteOrder"><span class="Definition">byte
order</span></a> in which the most significant byte comes first,
then the less significant bytes in descending order of
significance (<a href="#3MSB"><span class=
"Definition">MSB</span></a> <a href="#3LSB"><span class=
"Definition">LSB</span></a> for two-byte integers, <a href=
"#3MSB"><span class="Definition">MSB</span></a> B2 B1 <a href=
"#3LSB"><span class="Definition">LSB</span></a> for four-byte
integers).</dd>

<dt><a name="3palette">3.1.31 palette</a></dt>

<dd>indexed table of three 8-bit <a href="#3sample"><span class=
"Definition">sample</span></a> values, red, green, and blue,
which with an <a href="#3indexedColour"><span class=
"Definition">indexed-colour</span></a> image defines the red,
green, and blue sample values of the <a href=
"#3referenceImage"><span class="Definition">reference
image</span></a>. In other cases, the palette may be a suggested
palette that viewers may use to present the image on
indexed-colour display hardware. <a href="#3alpha"><span class=
"Definition">Alpha</span></a> samples may be defined for palette
entries via the <a href="#3alphaTable"><span class=
"Definition">alpha table</span></a> and may be used to
reconstruct the alpha sample values of the reference image.</dd>

<dt><a name="3passExtraction">3.1.32 pass extraction</a></dt>

<dd>organizing a <a href="#3PNGimage"><span class=
"Definition">PNG image</span></a> as a sequence of <a href=
"#3reducedImage"><span class="Definition">reduced
images</span></a> to change the order of transmission and enable
progressive display.</dd>

<dt><a name="3pixel">3.1.33 pixel</a></dt>

<dd>information stored for a single grid point in an image. A
pixel consists of (or points to) a sequence of <a href="#3sample"><span class=
"Definition">samples</span></a> from all <a href=
"#3channel"><span class="Definition">channels</span></a>. The
complete image is a rectangular array of pixels.</dd>


<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->

<dt><a name="3PNGdatastream">3.1.34 PNG datastream</a></dt>

<dd>result of encoding a <a href="#3PNGimage"><span class=
"Definition">PNG image</span></a>. A PNG <a href=
"#3datastream"><span class="Definition">datastream</span></a>
consists of a <a href="#3PNGsignature"><span class=
"Definition">PNG signature</span></a> followed by a sequence of
<a href="#3chunk"><span class=
"Definition">chunks</span></a>.</dd>

<dt><a name="3PNGdecoder">3.1.35 PNG decoder</a></dt>

<dd>process or device which reconstructs the <a href=
"#3referenceImage"><span class="Definition">reference
image</span></a> from a <a href="#3PNGdatastream"><span class=
"Definition">PNG datastream</span></a> and generates a
corresponding delivered image.</dd>

<dt><a name="3PNGeditor">3.1.36 PNG editor</a></dt>

<dd>process or device which creates a modification of an existing
<a href="#3PNGdatastream"><span class="Definition">PNG
datastream</span></a>, preserving unmodified ancillary
information wherever possible, and obeying the <a href=
"#3chunk"><span class="Definition">chunk</span></a> ordering
rules, even for unknown chunk types.</dd>

<dt><a name="3PNGencoder">3.1.37 PNG encoder</a></dt>

<dd>process or device which constructs a <a href=
"#3referenceImage"><span class="Definition">reference
image</span></a> from a <a href="#3sourceImage"><span class=
"Definition">source image</span></a>, and generates a <a href=
"#3PNGdatastream"><span class="Definition">PNG
datastream</span></a> representing the reference image.</dd>

<dt><a name="3PNGfile">3.1.38 PNG file</a></dt>

<dd><a href="#3PNGdatastream"><span class="Definition">PNG
datastream</span></a> stored as a file.</dd>

<dt><a name="3PNGfourByteSignedInteger">3.1.39 PNG four-byte
signed integer</a></dt>

<dd>a four-byte signed integer limited to the range
-(2<sup>31</sup>-1) to 2<sup>31</sup>-1. The restriction is
imposed in order to accommodate languages that have difficulty
with the value -2<sup>31</sup>.</dd>

<dt><a name="3PNGfourByteUnSignedInteger">3.1.40 PNG four-byte
unsigned integer</a></dt>

<dd>a four-byte unsigned integer limited to the range 0 to
2<sup>31</sup>-1. The restriction is imposed in order to
accommodate languages that have difficulty with unsigned
four-byte values.</dd>

<dt><a name="3PNGimage">3.1.41 PNG image</a></dt>

<dd>result of transformations applied by a <a href=
"#3PNGencoder"><span class="Definition">PNG encoder</span></a> to
a <a href="#3referenceImage"><span class="Definition">reference
image</span></a>, in preparation for encoding as a <a href=
"#3PNGdatastream"><span class="Definition">PNG
datastream</span></a>, and the result of decoding a PNG
datastream.</dd>

<dt><a name="3PNGsignature">3.1.42 PNG signature</a></dt>

<dd>sequence of <a href="#3byte"><span class=
"Definition">bytes</span></a> appearing at the start of every <a
href="#3PNGdatastream"><span class="Definition">PNG
datastream</span></a>. It differentiates a PNG datastream from
other types of <a href="#3datastream"><span class=
"Definition">datastream</span></a> and allows early detection of
some transmission errors.</dd>

<dt><a name="3reducedImage">3.1.43 reduced image</a></dt>

<dd>pass of the <a href="#3interlacedPNGimage"><span class=
"Definition">interlaced PNG image</span></a> extracted from the
<a href="#3PNGimage"><span class="Definition">PNG
image</span></a> by <a href="#3passExtraction"><span class=
"Definition">pass extraction</span></a>.</dd>

<dt><a name="3referenceImage">3.1.44 reference image</a></dt>

<dd>rectangular array of rectangular <a href="#3pixel"><span
class="Definition">pixels</span></a>, each having the same number
of <a href="#3sample"><span class=
"Definition">samples</span></a>, either three (red, green, blue)
or four (red, green, blue, <a href="#3alpha"><span class=
"Definition">alpha</span></a>). Every reference image can be
represented exactly by a <a href="#3PNGdatastream"><span class=
"Definition">PNG datastream</span></a> and every PNG datastream
can be converted into a reference image. Each <a href=
"#3channel"><span class="Definition">channel</span></a> has a <a
href="#3sampleDepth"><span class="Definition">sample
depth</span></a> in the range 1 to 16. All samples in the same
channel have the same sample depth. Different channels may have
different sample depths.</dd>

<dt><a name="3RGBmerging">3.1.45 RGB merging</a></dt>

<dd>converting an image in which the red, green, and blue <a
href="#3sample"><span class="Definition">samples</span></a> for
each <a href="#3pixel"><span class="Definition">pixel</span></a>
have the same value, and the same <a href="#3sampleDepth"><span
class="Definition">sample depth</span></a>, into an image with a
single <a href="#3greyscale"><span class=
"Definition">greyscale</span></a> <a href="#3channel"><span
class="Definition">channel</span></a>.</dd>

<dt><a name="3sample">3.1.46 sample</a></dt>

<dd>intersection of a <a href="#3channel"><span class=
"Definition">channel</span></a> and a <a href="#3pixel"><span
class="Definition">pixel</span></a> in an image.</dd>

<dt><a name="3sampleDepth">3.1.47 sample depth</a></dt>

<dd>number of bits used to represent a <a href="#3sample"><span
class="Definition">sample</span></a> value. In an <a href=
"#3indexedColour"><span class=
"Definition">indexed-colour</span></a> <a href="#3PNGimage"><span
class="Definition">PNG image</span></a>, samples are stored in
the <a href="#3palette"><span class=
"Definition">palette</span></a> and thus the sample depth is
always 8 by definition of the palette. In other types of PNG
image it is the same as the <a href="#3bitDepth"><span class=
"Definition">bit depth</span></a>.</dd>

<dt><a name="3sampleDepthScaling">3.1.48 sample depth
scaling</a></dt>

<dd>mapping of a range of <a href="#3sample"><span class=
"Definition">sample</span></a> values onto the full range of a <a
href="#3sampleDepth"><span class="Definition">sample
depth</span></a> allowed in a <a href="#3PNGimage"><span class=
"Definition">PNG image</span></a>.</dd>

<dt><a name="3scanline">3.1.49 scanline</a></dt>

<dd>row of <a href="#3pixel"><span class=
"Definition">pixels</span></a> within an image or <a href=
"#3interlacedPNGimage"><span class="Definition">interlaced PNG
image</span></a>.</dd>

<dt><a name="3sourceImage">3.1.50 source image</a></dt>

<dd>image which is presented to a <a href="#3PNGencoder"><span
class="Definition">PNG encoder</span></a>.</dd>

<dt><a name="3truecolour">3.1.51 truecolour</a></dt>

<dd>image representation in which each <a href="#3pixel"><span
class="Definition">pixel</span></a> is defined by <a href=
"#3sample"><span class="Definition">samples</span></a>,
representing red, green, and blue intensities and optionally an
<a href="#3alpha"><span class="Definition">alpha</span></a>
sample (in which case it is referred to as truecolour with
alpha).</dd>

<dt><a name="3whitePoint">3.1.52 white point</a></dt>

<dd><a href="#3chromaticity"><span class=
"Definition">chromaticity</span></a> of a computer display's
nominal white value.</dd>

<dt><a name="3zlib">3.1.53 zlib</a></dt>

<dd>particular format for data that have been compressed using <a
href="#3deflate"><span class=
"Definition">deflate</span></a>-style compression. Also the name
of a library containing a sample implementation of this method.
The format is defined in <a href="#2-RFC-1950"><span class=
"NormRef">[RFC-1950]</span></a>.</dd>
</dl>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="3Abbreviations">3.2 Abbreviated terms</a></h2>

<dl>
<dt><a name="3CRC">3.2.1 CRC</a></dt>

<dd>Cyclic Redundancy Code. A CRC is a type of check value
designed to detect most transmission errors. A decoder calculates
the CRC for the received data and checks by comparing it to the
CRC calculated by the encoder and appended to the data.
A mismatch
indicates that the data or the CRC were corrupted in
transit.</dd>

<dt><a name="3CRT">3.2.2 CRT</a></dt>

<dd>Cathode Ray Tube: a common type of computer display
hardware.</dd>

<dt><a name="3LSB">3.2.2 LSB</a></dt>

<dd>Least Significant Byte of a multi-<a href="#3byte"><span
class="Definition">byte</span></a> value.</dd>

<dt><a name="3LUT">3.2.3 LUT</a></dt>

<dd>Look Up Table. In <a href="#3frameBuffer"><span class=
"Definition">frame buffer</span></a> hardware, a LUT can be used
to map <a href="#3indexedColour"><span class=
"Definition">indexed-colour</span></a> <a href="#3pixel"><span
class="Definition">pixels</span></a> into a selected set of <a
href="#3truecolour"><span class=
"Definition">truecolour</span></a> values, or to perform <a href=
"#3gamma"><span class="Definition">gamma</span></a> correction.
In software, a LUT can often be used as a fast way of
implementing any mathematical function of a single integer
variable.</dd>

<dt><a name="3MSB">3.2.4 MSB</a></dt>

<dd>Most Significant Byte of a multi-<a href="#3byte"><span
class="Definition">byte</span></a> value.</dd>
</dl>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="4Concepts">4 Concepts</a></h1>

<h2><a name="4Concepts.Sourceimage">4.1 Images</a></h2>

<p>This International Standard specifies the PNG datastream, and
places some requirements on PNG encoders, which generate PNG
datastreams, PNG decoders, which interpret PNG datastreams, and
PNG editors, which transform one PNG datastream into another. It
does not specify the interface between an application and either
a PNG encoder, decoder, or editor. The precise form in which an
image is presented to an encoder or delivered by a decoder is not
specified. Four kinds of image are distinguished.</p>

<ol>
<li>The <i>source image</i> is the image presented to a PNG
encoder.</li>

<li>The <i>reference image</i>, which only exists conceptually,
is a rectangular array of rectangular pixels, all having the same
width and height, and all containing the same number of unsigned
integer samples, either three (red, green, blue) or four (red,
green, blue, alpha). The array of all samples of a particular
kind (red, green, blue, or alpha) is called a channel. Each
channel has a sample depth in the range 1 to 16, which is the
number of bits used by every sample in the channel. Different
channels may have different sample depths. The red, green, and
blue samples determine the intensities of the red, green, and
blue components of the pixel's colour; if they are all zero, the
pixel is black, and if they all have their maximum values
(2<sup>sampledepth</sup>-1), the pixel is white. The alpha sample
determines a pixel's degree of opacity, where zero means fully
transparent and the maximum value means fully opaque. In a
three-channel reference image all pixels are fully opaque. (It is
also possible for a four-channel reference image to have all
pixels fully opaque; the difference is that the latter has a
specific alpha sample depth, whereas the former does not.) Each
horizontal row of pixels is called a scanline. Pixels are ordered
from left to right within each scanline, and scanlines are
ordered from top to bottom. A PNG encoder may transform the source
image directly into a PNG image, but conceptually it first
transforms the source image into a reference image, then
transforms the reference image into a PNG image. Depending on the
type of source image, the transformation from the source image to
a reference image may require the loss of information. That
transformation is beyond the scope of this International
Standard. The reference image, however, can always be recovered
exactly from a PNG datastream.</li>

<li>The <i>PNG image</i> is obtained from the reference image by
a series of transformations: alpha separation, indexing, RGB
merging, alpha compaction, and sample depth scaling. Five types
of PNG image are defined (see 6.1: <a href=
"#6Colour-values"><span class="xref">Colour types and
values</span></a>). (If the PNG encoder actually transforms the
source image directly into the PNG image, and the source image
format is already similar to the PNG image format, the encoder
may be able to avoid doing some of these transformations.)
Although not all sample depths in the range 1 to 16 bits are
explicitly supported in the PNG image, the number of significant
bits in each channel of the reference image may be recorded. All
channels in the PNG image have the same sample depth. A PNG
encoder generates a PNG datastream from the PNG image. A PNG
decoder takes the PNG datastream and recreates the PNG
image.</li>

<li>The <i>delivered image</i> is constructed from the PNG image
obtained by decoding a PNG datastream. No specific format is
specified for the delivered image. A viewer presents an image to
the user as close to the appearance of the original source image
as it can achieve.</li>
</ol>

<p>The relationships between the four kinds of image are
illustrated in <a href="#figure41"><span class="figref">figure
4.1</span></a>.</p>

<p><a name="figure41">
<object data="figures/fig41.svg" type="image/svg+xml" width="640" height="290">
   <img height="280" width="640" src="png-figures/fig41.png" alt="Figure 4.1: Relationships between
source, reference, PNG, and display images" />
</object>
</a></p>

<p class="Figuretitle">Figure 4.1 &mdash; Relationships between
source, reference, PNG, and display images</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p>The relationships between samples, channels, pixels, and
sample depth are illustrated in <a href="#figure42"><span class=
"figref">figure 4.2</span></a>.</p>

<p><a name="figure42">
<object data="figures/fig42.svg" type="image/svg+xml" width="640" height="290">
  <img height="290" width="640" src="png-figures/fig42.png" alt="Figure 4.2: Relationships between
sample, sample depth, pixel, and channel" />
</object>
</a></p>

<p class="Figuretitle">Figure 4.2 &mdash; Relationships between
sample, sample depth, pixel, and channel</p>

<h2><a name="4Concepts.ColourSpaces">4.2 Colour spaces</a></h2>

<p>The RGB colour space in which colour samples are situated may
be specified in one of three ways:</p>

<!-- <ol start="1"> --><ol>
<li>by an ICC profile;</li>

<li>by specifying explicitly that the colour space is sRGB when
the samples conform to this colour space;</li>

<li>by specifying the value of gamma and the 1931 CIE <i>x,y</i>
chromaticities of the red, green, and blue primaries used in the
image and the reference white point.</li>
</ol>

<p>For high-end applications the first method provides the most
flexibility and control. The second method enables one particular
colour space to be indicated. The third method enables the exact
chromaticities of the RGB data to be specified, along with the
gamma correction (the power function relating the desired display
output with the image samples) to be applied (see Annex C: <a
href="#C-GammaAppendix"><span class="xref">Gamma and
chromaticity</span></a>). It is recommended that explicit gamma
information also be provided when either the first or second
method is used, for use by PNG decoders that do not support full
ICC profiles or the sRGB colour space. Such PNG decoders can
still make sensible use of gamma information. PNG decoders are
strongly encouraged to use this information, plus information
about the display system, in order to present the image to the
viewer in a way that reproduces as closely as possible what the image's original author
saw .</p>

<p>Gamma correction is not applied to the alpha channel, if
present. Alpha samples always represent a linear fraction of full
opacity.</p>

<h2><a name="4Concepts.PNGImageTransformation">4.3 Reference
image to PNG image transformation</a></h2>

<h3><a name="4Concepts.Introduction">4.3.1 Introduction</a></h3>

<p>A number of transformations are applied to the reference image
to create the PNG image to be encoded (see <a href=
"#figure43"><span class="figref">figure 4.3</span></a>). The
transformations are applied in the following sequence, where
square brackets mean the transformation is optional:</p>

<pre>
        [alpha separation]
        indexing or ( [RGB merging] [alpha compaction] )
        sample depth scaling
</pre>

<p>When every pixel is either fully transparent or fully opaque,
the alpha separation, alpha compaction, and indexing
transformations can cause the recovered reference image to have
an alpha sample depth different from the original reference
image, or to have no alpha channel. This has no effect on the
degree of opacity of any pixel. The two reference images are
considered equivalent, and the transformations are considered
lossless. Encoders that nevertheless wish to preserve the alpha
sample depth may elect not to perform transformations that would
alter the alpha sample depth.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p><a name="figure43">
<object data="figures/fig43.svg" type="image/svg+xml" height="525" width="640">
<img height="525" width="640" src="png-figures/fig43.png" alt="Figure 4.3: Reference image to PNG
image transformation" />
</object>
</a></p>

<p class="Figuretitle">Figure 4.3 &mdash; Reference image to PNG
image transformation</p>

<h3><a name="4Concepts.Implied-alpha">4.3.2 Alpha
separation</a></h3>

<p>If all alpha samples in a reference image have the maximum
value, then the alpha channel may be omitted, resulting in an
equivalent image that can be encoded more compactly.</p>

<h3><a name="4Concepts.Indexing">4.3.3 Indexing</a></h3>

<p>If the number of distinct pixel values is 256 or less, and the
RGB sample depths are not greater than 8, and the alpha channel
is absent or exactly 8 bits deep or every pixel is either fully
transparent or fully opaque, then an alternative representation
called indexed-colour may be more efficient for encoding.
Each pixel is replaced by an index into a palette.
The palette is a list of entries each containing
three 8-bit samples (red, green, blue). If an alpha channel is
present, there is also a parallel table of 8-bit alpha
samples.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p><a name="figure44">
<object height="450" width="660" data="figures/fig44.svg" type="image/svg+xml">
  <img height="450" width="660" src="png-figures/fig44.png" alt="Figure 4.4: Indexed-colour
image" />
</object>
</a></p>

<p class="Figuretitle">Figure 4.4 &mdash; Indexed-colour
image</p>

<p>A suggested palette or palettes may be constructed even when
the PNG image is not indexed-colour in order to assist viewers
that are capable of displaying only a limited number of
colours.</p>

<p>For indexed-colour images, encoders can rearrange the palette
so that the table entries with the maximum alpha value are
grouped at the end. In this case the table can be encoded in a
shortened form that does not include these entries.</p>

<h3><a name="4Concepts.RGBMerging">4.3.4 RGB merging</a></h3>

<p>If the red, green, and blue channels have the same sample
depth, and for each pixel the values of the red, green, and blue
samples are equal, then these three channels may be merged into a
single greyscale channel.</p>

<h3><a name="4Concepts.Alpha-indexing">4.3.5 Alpha
compaction</a></h3>

<p>For non-indexed images, if there exists an RGB (or greyscale)
value such that all pixels with that value are fully transparent
while all other pixels are fully opaque, then the alpha channel
can be represented more compactly by merely identifying the RGB
(or greyscale) value that is transparent.</p>

<h3><a name="4Concepts.Scaling">4.3.6 Sample depth
scaling</a></h3>

<p>In the PNG image, not all sample depths are supported (see
6.1: <a href="#6Colour-values"><span class="xref">Colour types
and values</span></a>), and all channels shall have the same
sample depth. All channels of the PNG image use the smallest
allowable sample depth that is not less than any sample depth in
the reference image, and the possible sample values in the
reference image are linearly mapped into the next allowable range
for the PNG image. <a href="#figure45"><span class=
"figref">Figure 4.5</span></a> shows how samples of depth 3 might
be mapped into samples of depth 4.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p><a name="figure45">
<object height="320" width="640" data="figures/fig45.svg" type="image/svg+xml">
  <img height="320" width="640" src="png-figures/fig45.png" alt="Figure 4.5: Scaling sample
values" />
</object>
</a></p>

<p class="Figuretitle">Figure 4.5 &mdash; Scaling sample
values</p>

<p>Allowing only a few sample depths reduces the number of cases
that decoders have to cope with. Sample depth scaling is
reversible with no loss of data, because the reference image
sample depths can be recorded in the PNG datastream. In the
absence of recorded sample depths, the reference image sample
depth equals the PNG image sample depth. See 12.5: <a href=
"#12Sample-depth-scaling"><span class="xref">Sample depth
scaling</span></a> and 13.12: <a href=
"#13Sample-depth-rescaling"><span class="xref">Sample depth
rescaling</span></a>.</p>

<p><a name="figure46">
<object height="450" width="660" data="figures/fig46.svg" type="image/svg+xml">
  <img  height="450" width="660" src= "png-figures/fig46.png" alt="Figure 4.6: Possible PNG image
pixel types" />
</object>
</a></p>

<p class="Figuretitle">Figure 4.6 &mdash; Possible PNG image
pixel types</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="4Concepts.PNGImage">4.4 PNG image</a></h2>

<p>The transformation of the reference image results in one of
five types of PNG image (see <a href="#figure46"><span class=
"figref">figure 4.6</span></a>) :</p>

<!-- <ol start="1"> --><ol>
<li>Truecolour with alpha: each pixel consists of four samples:
red, green, blue, and alpha.</li>

<li>Greyscale with alpha: each pixel consists of two samples:
grey and alpha.</li>

<li>Truecolour: each pixel consists of three samples: red, green,
and blue. The alpha channel may be represented by a single pixel
value. Matching pixels are fully transparent, and all others are
fully opaque. If the alpha channel is not represented in this
way, all pixels are fully opaque.</li>

<li>Greyscale: each pixel consists of a single sample: grey. The
alpha channel may be represented by a single pixel value as in
the previous case. If the alpha channel is not represented in
this way, all pixels are fully opaque.</li>

<li>Indexed-colour: each pixel consists of an index into a palette (and into an associated table of alpha values, if present).</li>
</ol>

<p>The format of each pixel depends on the PNG image type and the
bit depth. For PNG image types other than indexed-colour,
the bit depth specifies the number of bits per sample, not the
total number of bits per pixel.
For indexed-colour images, the bit depth specifies the
number of bits in each palette index, not the sample depth of the
colours in the palette or alpha table. Within the pixel the
samples appear in the following order, depending on the PNG image
type.</p>

<!-- <ol start="6"> --><ol>
<li>Truecolour with alpha: red, green, blue, alpha.</li>

<li>Greyscale with alpha: grey, alpha.</li>

<li>Truecolour: red, green, blue.</li>

<li>Greyscale: grey.</li>

<li>Indexed-colour: palette index.</li>
</ol>

<h2><a name="4Concepts.Encoding">4.5 Encoding the PNG
image</a></h2>

<h3><a name="4Concepts.EncodingIntro">4.5.1 Introduction</a></h3>

<p>A conceptual model of the process of encoding a PNG image is
given in <a href="#figure47"><span class="figref">figure
4.7</span></a>. The steps refer to the operations on the array of
pixels or indices in the PNG image. The palette and alpha table
are not encoded in this way.</p>

<!-- <ol start="1"> --><ol>
<li>Pass extraction: to allow for progressive display, the PNG
image pixels can be rearranged to form several smaller images
called reduced images or passes.</li>

<li>Scanline serialization: the image is serialized a scanline at
a time. Pixels are ordered left to right in a scanline and
scanlines are ordered top to bottom.</li>

<li>Filtering: each scanline is transformed into a filtered
scanline using one of the defined filter types to prepare the
scanline for image compression.</li>

<li>Compression: occurs on all the filtered scanlines in the
image.</li>

<li>Chunking: the compressed image is divided into conveniently
sized chunks. An error detection code is added to each
chunk.</li>

<li>Datastream construction: the chunks are inserted into the
datastream.</li>
</ol>

<h3><a name="4Concepts.EncodingPassAbs">4.5.2 Pass
extraction</a></h3>

<p>Pass extraction (see <a href="#figure48"><span class=
"figref">figure 4.8</span></a>) splits a PNG image into a
sequence of reduced images where the first image defines a coarse
view and subsequent images enhance this coarse view until the
last image completes the PNG image. The set of reduced images is
also called an interlaced PNG image. Two interlace methods are
defined in this International Standard. The first method is a
null method; pixels are stored sequentially from left to right
and scanlines from top to bottom. The second method makes
multiple scans over the image to produce a sequence of seven
reduced images. The seven passes for a sample image are
illustrated in <a href="#figure48"><span class="figref">figure
4.8</span></a>. See clause&#160;8: <a href="#8Interlace"><span class=
"xref">Interlacing and pass extraction</span></a>.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p><a name="figure47">
<object height="575" width="645" data="figures/fig47.svg" type="image/svg+xml">
	<img height="575" width="645" src="png-figures/fig47.png" alt="Figure 4.7: Encoding the PNG
image" />
</object>
</a></p>

<p class="Figuretitle">Figure 4.7 &mdash; Encoding the PNG
image</p>

<p><a name="figure48">
<object height="450" width="645" data="figures/fig48.svg" type="image/svg+xml">
	<img height="450" width="645" src="png-figures/fig48.png" alt="Figure 4.8: Pass extraction" />
</object>
</a></p>

<p class="Figuretitle">Figure 4.8 &mdash; Pass extraction</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h3><a name="4Concepts.EncodingScanlineAbs">4.5.3 Scanline
serialization</a></h3>

<p>Each row of pixels, called a scanline, is represented as a
sequence of bytes.</p>

<h3><a name="4Concepts.EncodingFiltering">4.5.4
Filtering</a></h3>

<p>PNG standardizes one filter method and several filter types
that may be used to prepare image data for compression. It
transforms the byte sequence in a scanline to an equal length
sequence of bytes preceded by a filter type byte (see <a href=
"#figure49"><span class="figref">figure 4.9</span></a> for an
example). The filter type byte defines
the specific filtering to be applied to a specific
scanline. The encoder shall use only a single filter method for
an interlaced PNG image, but may use different filter types for
each scanline in a reduced image. See clause&#160;9: <a href=
"#9Filters"><span class="xref">Filtering</span></a>.</p>

<p><a name="figure49">
<object height="340" width="710" data="figures/fig49.svg" type="image/svg+xml">
  <img height="340" width="710" src="png-figures/fig49.png" alt="Figure 4.9: Serializing and
filtering a scanline" />
</object>
</a></p>

<p class="Figuretitle">Figure 4.9 &mdash; Serializing and
filtering a scanline</p>

<h3><a name="4Concepts.EncodingCompression">4.5.5
Compression</a></h3>

<p>The sequence of filtered scanlines in the pass or passes of
the PNG image is compressed (see <a href="#figure410"><span
class="figref">figure 4.10</span></a>) by one of the defined
compression methods. The concatenated filtered scanlines form the
input to the compression stage. The output from the compression
stage is a single compressed datastream. See clause&#160;10: <a href=
"#10Compression"><span class="xref">Compression</span></a>.</p>

<h3><a name="4Concepts.EncodingChunking">4.5.6 Chunking</a></h3>

<p>Chunking provides a convenient breakdown of the compressed
datastream into manageable chunks (see <span
class="figref"><a href="#figure410">figure 4.10</a></span>). Each chunk has its own
redundancy check. See clause&#160;11: <a href="#11Chunks"><span class=
"xref">Chunk specifications</span></a>.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p><a name="figure410">
<object height="450" width="700" data="figures/fig410.svg" type="image/svg+xml">
 <img height="450" width="700" src="png-figures/fig410.png" alt="Figure 4.10: Compression" />
</object></a></p>
<p class="Figuretitle">Figure 4.10 &mdash; Compression</p>

<h2><a name="4Concepts.AncillInfo">4.6 Additional
information</a></h2>

<p>Ancillary information may be associated with an image.
Decoders may ignore all or some of the ancillary information. The
types of ancillary information provided are described in <a href=
"#table41"><span class="tabref">Table 4.1</span></a>.</p>

<table class="Regular" summary=
"This table lists the types of ancillary information that may be associated with an image">
<caption><a name="table41"><b>Table 4.1 &mdash; Types of
ancillary information</b></a></caption>

<tr>
<th>Type of information</th>
<th>Description</th>
</tr>

<tr>
<td class="Regular">Background colour</td>
<td class="Regular">Solid background colour to be used when presenting the image
if no better option is available.</td>
</tr>

<tr>
<td class="Regular">Gamma and chromaticity</td>
<td class="Regular">Gamma characteristic of the image with respect to the desired
output intensity, and chromaticity characteristics of the RGB
values used in the image.</td>
</tr>

<tr>
<td class="Regular">ICC profile</td>
<td class="Regular">Description of the colour space (in the form of an
International Color Consortium (ICC) profile) to which the
samples in the image conform.</td>
</tr>

<tr>
<td class="Regular">Image histogram</td>
<td class="Regular">Estimates of how frequently the image uses each palette entry.</td>
</tr>

<tr>
<td class="Regular">Physical pixel dimensions</td>
<td class="Regular">Intended pixel size and aspect ratio to be used in presenting
the PNG image.</td>
</tr>

<tr>
<td class="Regular">Significant bits</td>
<td class="Regular">The number of bits that are significant in the samples.</td>
</tr>

<tr>
<td class="Regular">sRGB colour space</td>
<td class="Regular">A rendering intent (as defined by the International Color
Consortium) and an indication that the image samples conform to
this colour space.</td>
</tr>

<tr>
<td class="Regular">Suggested palette</td>
<td class="Regular">A reduced palette that may be used when the display device is
not capable of displaying the full range of colours in the
image.</td>
</tr>

<tr>
<td class="Regular">Textual data</td>
<td class="Regular">Textual information (which may be compressed) associated with
the image.</td>
</tr>

<tr>
<td class="Regular">Time</td>
<td class="Regular">The time when the PNG image was last modified.</td>
</tr>

<tr>
<td class="Regular">Transparency</td>
<td class="Regular">Alpha information that allows the reference image to be
reconstructed when the alpha channel is not retained in the PNG
image.</td>
</tr>
</table>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="4Concepts.Format">4.7 PNG datastream</a></h2>

<h3><a name="4Concepts.FormatChunks">4.7.1 Chunks</a></h3>

<p>The PNG datastream consists of a PNG signature (see 5.2: <a
href="#5PNG-file-signature"><span class="xref">PNG
signature</span></a>) followed by a sequence of chunks (see
clause&#160;11: <a href="#11Chunks"><span class="xref">Chunk
specifications</span></a>). Each chunk has a chunk type which
specifies its function.</p>

<h3><a name="4Concepts.FormatTypes">4.7.2 Chunk types</a></h3>

<p>There are 18 chunk types defined in this International
Standard. Chunk types are four-byte sequences chosen so that they
correspond to readable labels when interpreted in the ISO 646.IRV:1991
character set. The first four are termed critical chunks, which
shall be understood and correctly interpreted according to the
provisions of this International Standard. These are:</p>

<!-- <ol start="1"> --><ol>
<li><a href="#11IHDR"><span class="chunk">IHDR</span></a>: image
header, which is the first chunk in a PNG datastream.</li>

<li><a href="#11PLTE"><span class="chunk">PLTE</span></a>:
palette table associated with indexed PNG images.</li>

<li><a href="#11IDAT"><span class="chunk">IDAT</span></a>: image
data chunks.</li>

<li><a href="#11IEND"><span class="chunk">IEND</span></a>: image
trailer, which is the last chunk in a PNG datastream.</li>
</ol>

<p>The remaining 14 chunk types are termed ancillary chunk types,
which encoders may generate and decoders may interpret.</p>

<!-- <ol start="5"> --><ol>
<li>Transparency information: <a href="#11tRNS"><span class=
"chunk">tRNS</span></a> (see 11.3.2: <a class='Href' href=
'#11transinfo'>Transparency information</a>).</li>

<li>Colour space information: <a href="#11cHRM"><span class=
"chunk">cHRM</span></a>, <a href="#11gAMA"><span class=
"chunk">gAMA</span></a>, <a href="#11iCCP"><span class=
"chunk">iCCP</span></a>, <a href="#11sBIT"><span class=
"chunk">sBIT</span></a>, <a href="#11sRGB"><span class=
"chunk">sRGB</span></a> (see 11.3.3: <a class='Href' href=
'#11addnlcolinfo'>Colour space information</a>).</li>

<li>Textual information: <a href="#11iTXt"><span class=
"chunk">iTXt</span></a>, <a href="#11tEXt"><span class=
"chunk">tEXt</span></a>, <a href="#11zTXt"><span class=
"chunk">zTXt</span></a> (see 11.3.4: <a class='Href' href=
'#11textinfo'>Textual information</a>).</li>

<li>Miscellaneous information: <a href="#11bKGD"><span class=
"chunk">bKGD</span></a>, <a href="#11hIST"><span class=
"chunk">hIST</span></a>, <a href="#11pHYs"><span class=
"chunk">pHYs</span></a>, <a href="#11sPLT"><span class=
"chunk">sPLT</span></a> (see 11.3.5: <a class='Href' href=
'#11addnlsiinfo'>Miscellaneous information</a>).</li>

<li>Time information: <a href="#11tIME"><span class=
"chunk">tIME</span></a> (see 11.3.6: <a class='Href' href=
'#11timestampinfo'>Time stamp information</a>).</li>
</ol>

<h2><a name="4Concepts.Errors">4.8 Error handling</a></h2>

<p>Errors in a PNG datastream fall into two general classes:</p>

<!-- <ol start="1"> --><ol>
<li>transmission errors or damage to a computer file system,
which tend to corrupt much or all of the datastream;</li>

<li>syntax errors, which appear as invalid values in chunks, or
as missing or misplaced chunks. Syntax errors can be caused not
only by encoding mistakes, but also by the use of registered or
private values, if those values are unknown to the decoder.</li>
</ol>

<p>PNG decoders should detect errors as early as possible,
recover from errors whenever possible, and fail gracefully
otherwise. The error handling philosophy is described in detail
in 13.2: <a href="#13Decoders.Errors"><span class="xref">Error
handling</span></a>.</p>

<h2><a name="4Concepts.Registration">4.9 Extension and
registration</a></h2>

<p>For some facilities in PNG, there are a number of alternatives
defined, and this International Standard allows other
alternatives to be defined by registration. According to the
rules for the designation and operation of registration
authorities in the ISO/IEC Directives, the ISO and IEC Councils
have designated the following as the registration authority:</p>

<address>The World-Wide Web Consortium Host at ERCIM</address>

<address>The Registration Authority for PNG</address>

<address>INRIA- Sophia Antipolis</address>

<address>BP 93</address>

<address>06902 Sophia Antipolis Cedex</address>

<address>FRANCE</address>

<address>Email:png-group@w3.org</address>

<p>To ensure timely processing the Registration Authority should be contacted by email.</p>

<p>The following entities may be registered:</p>

<!-- <ol start="1"> --><ol>
<li>chunk type;</li>

<li>text keyword.</li>
</ol>

<p>The following entities are reserved for future
standardization:</p>

<!-- <ol start="4"> --><ol>
<li>undefined field values less than 128;</li>

<li>filter method;</li>

<li>filter type;</li>

<li>interlace method;</li>

<li>compression method.</li>
</ol>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="5DataRep">5 Datastream structure</a></h1>

<h2><a name="5Introduction">5.1 Introduction</a></h2>

<p>This clause defines the PNG signature and the basic properties
of chunks. Individual chunk types are discussed in clause&#160;11: <a
href="#11Chunks"><span class="xref">Chunk
specifications</span></a>.</p>

<h2><a name="5PNG-file-signature">5.2 PNG signature</a></h2>

<p>The first eight bytes of a PNG datastream always contain the
following (decimal) values:</p>

<pre>
   137 80 78 71 13 10 26 10
</pre>

<p>This signature indicates that the remainder of the datastream
contains a single PNG image, consisting of a series of chunks
beginning with an <a href="#11IHDR"><span class=
"chunk">IHDR</span></a> chunk and ending with an <a href=
"#11IEND"><span class="chunk">IEND</span></a> chunk.</p>

<h2><a name="5Chunk-layout">5.3 Chunk layout</a></h2>

<p>Each chunk consists of three or four fields (see figure 5.1).
The meaning of the fields is described in 
<a href="#table51"><span class="tabref">Table 5.1</span></a>.
The chunk data field may be empty.</p>

<p><a name="figure411">
<object height="160" width="480" data="figures/fig51.svg" type="image/svg+xml">
 <img height="160" width="480" src="png-figures/fig51.png" alt="Figure 5.1: Chunk parts" />
</object>
</a></p>

<p class="Figuretitle">Figure 5.1 &mdash; Chunk parts</p>

<table class="Regular" summary=
"This table defines the chunk fields">
<caption><a name="table51"><b>Table 5.1 &mdash; Chunk fields</b></a></caption>
<tr>
<td class="Regular">Length</td>
<td class="Regular">A four-byte unsigned integer giving the number of bytes in
the chunk's data field. The length counts <strong>only</strong>
the data field, <strong>not</strong> itself, the chunk type, or
the CRC. Zero is a valid length. Although encoders and decoders
should treat the length as unsigned, its value shall not exceed
2<sup>31</sup>-1 bytes.</td>
</tr>

<tr>
<td class="Regular">Chunk Type</td>
<td class="Regular">A sequence of four bytes defining the chunk type. Each byte
of a chunk type is restricted to the decimal values 65 to 90 and
97 to 122. These correspond to the uppercase and lowercase ISO
646 letters (<tt>A</tt>-<tt>Z</tt> and <tt>a</tt>-<tt>z</tt>)
respectively for convenience in description and examination of
PNG datastreams. Encoders and decoders shall treat the chunk
types as fixed binary values, not character strings. For example,
it would not be correct to represent the chunk type <a href=
"#11IDAT"><span class="chunk">IDAT</span></a> by the equivalents
of those letters in the UCS 2 character set. Additional naming
conventions for chunk types are discussed in 5.4: <a href=
"#5Chunk-naming-conventions"><span class="xref">Chunk naming
conventions</span></a>.</td>
</tr>

<tr>
<td class="Regular">Chunk Data</td>
<td class="Regular">The data bytes appropriate to the chunk type, if any. This
field can be of zero length.</td>
</tr>

<tr>
<td class="Regular">CRC</td>
<td class="Regular">A four-byte CRC (Cyclic Redundancy Code) calculated on the
preceding bytes in the chunk, including the chunk type field and
chunk data fields, but <strong>not</strong> including the length
field. The CRC can be used to check for corruption of the data.
The CRC is always present, even for chunks containing no data.
See 5.5: <a href="#5CRC-algorithm"><span class="xref">Cyclic
Redundancy Code algorithm</span></a>.</td>
</tr>
</table>

<p>The chunk data length may be any number of bytes up to the
maximum; therefore, implementors cannot assume that chunks are
aligned on any boundaries larger than bytes.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="5Chunk-naming-conventions">5.4 Chunk naming
conventions</a></h2>

<p>Chunk types are chosen to be meaningful names when the bytes
of the chunk type are interpreted as ISO 646 letters. Chunk types
are assigned so that a decoder can determine some properties of a
chunk even when the type is not recognized. These rules allow
safe, flexible extension of the PNG format, by allowing a PNG
decoder to decide what to do when it encounters an unknown chunk.
(The chunk types standardized in this International Standard are
defined in clause&#160;11: <a href="#11Chunks"><span class=
"xref">Chunk specifications</span></a>, and the way to add
non-standard chunks is defined in clause&#160;14: <a href=
"#14EditorsExt"><span class="xref">Editors and
extensions</span></a>.) The naming rules are normally of interest
only when the decoder does not recognize the chunk's type.</p>

<p>Four bits of the chunk type, the property bits, namely bit 5
(value 32) of each byte, are used to convey chunk properties.
This choice means that a human can read off the assigned
properties according to whether the letter corresponding to each
byte of the chunk type is uppercase (bit 5 is 0) or lowercase
(bit 5 is 1). However, decoders should test the properties of an
unknown chunk type by numerically testing the specified bits;
testing whether a character is uppercase or lowercase is
inefficient, and even incorrect if a locale-specific case
definition is used.</p>

<p>The property bits are an inherent part of the chunk type, and
hence are fixed for any chunk type. Thus, <span class=
"chunk">CHNK</span> and <span class="chunk">cHNk</span> would be
unrelated chunk types, not the same chunk with different
properties.</p>

<p>The semantics of the property bits are
defined in
<a href="#table52"><span class="tabref">Table 5.2</span></a>.
</p>

<table class="Regular" summary=
"This table defines the semantics of the property bits">
<caption><a name="table52"><b>Table 5.2 &mdash; Semantics of property bits</b></a></caption>
<tr>
<td class="Regular">Ancillary bit: first byte</td>
<td class="Regular">0 (uppercase) = critical,<br class="xhtml" />
 1 (lowercase) = ancillary.</td>
<td class="Regular">Critical chunks are necessary for successful display of the
contents of the datastream, for example the image header chunk
(<a href="#11IHDR"><span class="chunk">IHDR</span></a>). A
decoder trying to extract the image, upon encountering an unknown
chunk type in which the ancillary bit is 0, shall indicate to the
user that the image contains information it cannot safely
interpret.<br class="xhtml" />
 Ancillary chunks are not strictly necessary in order to
meaningfully display the contents of the datastream, for example
the time chunk (<a href="#11tIME"><span class=
"chunk">tIME</span></a>). A decoder encountering an unknown chunk
type in which the ancillary bit is 1 can safely ignore the chunk
and proceed to display the image.</td>
</tr>

<tr>
<td class="Regular">Private bit: second byte</td>
<td class="Regular">0 (uppercase) = public,<br class="xhtml" />
 1 (lowercase) = private.</td>
<td class="Regular">A public chunk is one that is defined in this International
Standard or is registered in the list of PNG special-purpose
public chunk types maintained by the Registration Authority (see
4.9 <a href="#4Concepts.Registration"><span class=
"xref">Extension and registration</span></a>). Applications can
also define private (unregistered) chunk types for their own
purposes. The names of private chunks have a lowercase second
letter, while public chunks will always be assigned names with
uppercase second letters. Decoders do not need to test the
private-chunk property bit, since it has no functional
significance; it is simply an administrative convenience to
ensure that public and private chunk names will not conflict. See
clause&#160;14: <a href="#14EditorsExt"><span class="xref">Editors and
extensions</span></a> and 12.10.2: <a href=
"#12Use-of-private-chunks"><span class="xref">Use of private
chunks</span></a>.</td>
</tr>

<tr>
<td class="Regular">Reserved bit: third byte</td>
<td class="Regular">0 (uppercase) in this version of PNG.<br class="xhtml" />
 If the reserved bit is 1, the datastream does not conform to
this version of PNG.</td>
<td class="Regular">The significance of the case of the third letter of the chunk
name is reserved for possible future extension. In this
International Standard, all chunk names shall have uppercase
third letters.</td>
</tr>

<tr>
<td class="Regular">Safe-to-copy bit: fourth byte</td>
<td class="Regular">0 (uppercase) = unsafe to copy,<br class="xhtml" />
1 (lowercase) = safe to copy.</td>
<td class="Regular">This property bit is not of interest to pure decoders, but it
is needed by PNG editors. This bit defines the proper handling of
unrecognized chunks in a datastream that is being modified. Rules
for PNG editors are discussed further in 14.2: <a href=
"#14Ordering"><span class="xref">Behaviour of PNG
editors</span></a>.</td>
</tr>
</table>

<p>EXAMPLE The hypothetical chunk type "<span class=
"chunk">cHNk</span>" has the property bits:</p>

<pre>
   cHNk  &lt;-- 32 bit chunk type represented in text form
   ||||
   |||+- Safe-to-copy bit is 1 (lower case letter; bit 5 is 1)
   ||+-- Reserved bit is 0     (upper case letter; bit 5 is 0)
   |+--- Private bit is 0      (upper case letter; bit 5 is 0)
   +---- Ancillary bit is 1    (lower case letter; bit 5 is 1)
</pre>

<p>Therefore, this name represents an ancillary, public,
safe-to-copy chunk.</p>

<h2><a name="5CRC-algorithm">5.5 Cyclic Redundancy Code
algorithm</a></h2>

<p>CRC fields are calculated using standardized CRC methods with
pre and post conditioning, as defined by ISO 3309 <a href=
"#2-ISO-3309"><span class="NormRef">[ISO-3309]</span></a> and
ITU-T V.42 <a href="#G-ITU-T-V42"><span class=
"bibref">[ITU-T-V42]</span></a>. The CRC polynomial employed
is</p>

<p>x<sup>32</sup> + x<sup>26</sup> + x<sup>23</sup> +
x<sup>22</sup> + x<sup>16</sup> + x<sup>12</sup> + x<sup>11</sup>
+ x<sup>10</sup> + x<sup>8</sup> + x<sup>7</sup> + x<sup>5</sup>
+ x<sup>4</sup> + x<sup>2</sup> + x + 1</p>

<p>In PNG, the 32-bit CRC is initialized to all 1's, and then the
data from each byte is processed from the least significant bit
(1) to the most significant bit (128). After all the data bytes
are processed, the CRC is inverted (its ones complement is
taken). This value is transmitted (stored in the datastream) MSB
first. For the purpose of separating into bytes and ordering, the
least significant bit of the 32-bit CRC is defined to be the
coefficient of the <tt>x<sup>31</sup></tt> term.</p>

<p>Practical calculation of the CRC often employs a precalculated
table to accelerate the computation. See Annex D: <a href=
"#D-CRCAppendix"><span class="xref">Sample Cyclic Redundancy Code
implementation</span></a>.</p>

<h2><a name="5ChunkOrdering">5.6 Chunk ordering</a></h2>

<p>The constraints on the positioning of the individual chunks
are listed in <a href="#table53"><span class="tabref">Table
5.3</span></a> and illustrated diagrammatically in <a href=
"#figure52"><span class="figref">figure 5.2</span></a> and <a
href="#figure53"><span class="figref">figure 5.3</span></a>.
These lattice diagrams represent the constraints on positioning
imposed by this International Standard. The lines in the diagrams
define partial ordering relationships. Chunks higher up shall
appear before chunks lower down. Chunks which are horizontally
aligned and appear between two other chunk types (higher and
lower than the horizontally aligned chunks) may appear in any
order between the two higher and lower chunk types to which they
are connected. The superscript associated with the chunk type is
defined in <a href="#table54"><span class="tabref">Table
5.4</span></a>. It indicates whether the chunk is mandatory,
optional, or may appear more than once. A vertical bar between
two chunk types indicates alternatives.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<table class="Regular" summary=
"This table lists the chunk ordering rules">
<caption><a name="table53"><b>Table 5.3 &mdash; Chunk ordering
rules</b></a></caption>

<tr>
<th colspan="3">Critical chunks<br class="xhtml" />
 (shall appear in this order, except <a href="#11PLTE"><span
class="chunk">PLTE</span></a> is optional)</th>
</tr>

<tr>
<th>Chunk name</th>
<th>Multiple allowed</th>
<th>Ordering constraints</th>
</tr>

<tr>
<td class="Regular"><a href="#11IHDR"><span class="chunk">IHDR</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">Shall be first</td>
</tr>

<tr>
<td class="Regular"><a href="#11PLTE"><span class="chunk">PLTE</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">Before first <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> </td>
</tr>

<tr>
<td class="Regular"><a href="#11IDAT"><span class="chunk">IDAT</span></a> </td>
<td class="Regular">Yes</td>
<td class="Regular">Multiple <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunks shall be consecutive</td>
</tr>

<tr>
<td class="Regular"><a href="#11IEND"><span class="chunk">IEND</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">Shall be last</td>
</tr>

<tr>
<th colspan="3">Ancillary chunks<br class="xhtml" />
 (need not appear in this order)</th>
</tr>

<tr>
<th>Chunk name</th>
<th>Multiple allowed</th>
<th>Ordering constraints</th>
</tr>

<tr>
<td class="Regular"><a href="#11cHRM"><span class="chunk">cHRM</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">Before <a href="#11PLTE"><span class="chunk">PLTE</span></a>
and <a href="#11IDAT"><span class="chunk">IDAT</span></a> </td>
</tr>

<tr>
<td class="Regular"><a href="#11gAMA"><span class="chunk">gAMA</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">Before <a href="#11PLTE"><span class="chunk">PLTE</span></a>
and <a href="#11IDAT"><span class="chunk">IDAT</span></a> </td>
</tr>

<tr>
<td class="Regular"><a href="#11iCCP"><span class="chunk">iCCP</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">Before <a href="#11PLTE"><span class="chunk">PLTE</span></a>
and <a href="#11IDAT"><span class="chunk">IDAT</span></a>. If the
<a href="#11iCCP"><span class="chunk">iCCP</span></a> chunk is
present, the <a href="#11sRGB"><span class=
"chunk">sRGB</span></a> chunk should not be present.</td>
</tr>

<tr>
<td class="Regular"><a href="#11sBIT"><span class="chunk">sBIT</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">Before <a href="#11PLTE"><span class="chunk">PLTE</span></a>
and <a href="#11IDAT"><span class="chunk">IDAT</span></a> </td>
</tr>

<tr>
<td class="Regular"><a href="#11sRGB"><span class="chunk">sRGB</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">Before <a href="#11PLTE"><span class="chunk">PLTE</span></a>
and <a href="#11IDAT"><span class="chunk">IDAT</span></a>. If the
<a href="#11sRGB"><span class="chunk">sRGB</span></a> chunk is
present, the <a href="#11iCCP"><span class=
"chunk">iCCP</span></a> chunk should not be present.</td>
</tr>

<tr>
<td class="Regular"><a href="#11bKGD"><span class="chunk">bKGD</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">After <a href="#11PLTE"><span class="chunk">PLTE</span></a>;
before <a href="#11IDAT"><span class="chunk">IDAT</span></a>
</td>
</tr>

<tr>
<td class="Regular"><a href="#11hIST"><span class="chunk">hIST</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">After <a href="#11PLTE"><span class="chunk">PLTE</span></a>;
before <a href="#11IDAT"><span class="chunk">IDAT</span></a>
</td>
</tr>

<tr>
<td class="Regular"><a href="#11tRNS"><span class="chunk">tRNS</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">After <a href="#11PLTE"><span class="chunk">PLTE</span></a>;
before <a href="#11IDAT"><span class="chunk">IDAT</span></a>
</td>
</tr>

<tr>
<td class="Regular"><a href="#11pHYs"><span class="chunk">pHYs</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">Before <a href="#11IDAT"><span class="chunk">IDAT</span></a>
</td>
</tr>

<tr>
<td class="Regular"><a href="#11sPLT"><span class="chunk">sPLT</span></a> </td>
<td class="Regular">Yes</td>
<td class="Regular">Before <a href="#11IDAT"><span class="chunk">IDAT</span></a>
</td>
</tr>

<tr>
<td class="Regular"><a href="#11tIME"><span class="chunk">tIME</span></a> </td>
<td class="Regular">No</td>
<td class="Regular">None</td>
</tr>

<tr>
<td class="Regular"><a href="#11iTXt"><span class="chunk">iTXt</span></a> </td>
<td class="Regular">Yes</td>
<td class="Regular">None</td>
</tr>

<tr>
<td class="Regular"><a href="#11tEXt"><span class="chunk">tEXt</span></a> </td>
<td class="Regular">Yes</td>
<td class="Regular">None</td>
</tr>

<tr>
<td class="Regular"><a href="#11zTXt"><span class="chunk">zTXt</span></a> </td>
<td class="Regular">Yes</td>
<td class="Regular">None</td>
</tr>
</table>

<table class="Regular"  summary=
"This table lists the symbols used in lattice diagrams">
<caption><a name="table54"><b>Table 5.4 &mdash; Meaning of
symbols used in lattice diagrams</b></a></caption>

<tr>
<th>Symbol</th>
<th>Meaning</th>
</tr>

<tr>
<td class="Regular">+</td>
<td class="Regular">One or more</td>
</tr>

<tr>
<td class="Regular">1</td>
<td class="Regular">Only one</td>
</tr>

<tr>
<td class="Regular">?</td>
<td class="Regular">Zero or one</td>
</tr>

<tr>
<td class="Regular">*</td>
<td class="Regular">Zero or more</td>
</tr>
<tr>
<td class="Regular">|</td>
<td class="Regular">Alternative</td>
</tr>
</table>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p>
<object height="540" width="800" data="figures/fig52.svg" type="image/svg+xml">
 <img height="540" width="800" src="png-figures/fig52.png" alt="Figure 5.2: Lattice diagram: PNG images with PLTE in datastream" />
</object>
</p>

<p class="Figuretitle"><a name="figure52">Figure 5.2 &mdash;</a>
Lattice diagram: PNG images with <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> in datastream</p>

<p>
<object height="540" width="900" data="figures/fig53.svg"
type="image/svg+xml">
 <img height="540" width="900" src="png-figures/fig53.png" alt="Figure 5.3: Lattice diagram: PNG images without PLTE in datastream" />
</object>
</p>

<p class="Figuretitle"><a name="figure53">Figure 5.3 &mdash;</a>
Lattice diagram: PNG images without <a href="#11PLTE"><span
class="chunk">PLTE</span></a> in datastream</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="6Transformation">6 Reference image to PNG image
transformation</a></h1>

<h2><a name="6Colour-values">6.1 Colour types and values</a></h2>

<p>As explained in 4.4: <a href="#4Concepts.PNGImage"><span
class="xref">PNG image</span></a> there are five types of PNG
image. Corresponding to each type is a colour type, which is the
sum of the following values: 1 (palette used), 2 (truecolour
used) and 4 (alpha used). Greyscale and truecolour images may
have an explicit alpha channel. The PNG image types and
corresponding colour types are listed in <a href=
"#table6.1"><span class="tabref">Table 6.1</span></a>.</p>

<table class="Regular"  summary=
"This table lists the PNG image and colour types">
<caption><a name="table6.1"><b>Table 6.1 &mdash; PNG image types
and colour types</b></a></caption>

<tr>
<th>PNG image type</th>
<th>Colour type</th>
</tr>

<tr>
<td class="Regular">Greyscale</td>
<td class="Regular">0</td>
</tr>

<tr>
<td class="Regular">Truecolour</td>
<td class="Regular">2</td>
</tr>

<tr>
<td class="Regular">Indexed-colour</td>
<td class="Regular">3</td>
</tr>

<tr>
<td class="Regular">Greyscale with alpha</td>
<td class="Regular">4</td>
</tr>

<tr>
<td class="Regular">Truecolour with alpha</td>
<td class="Regular">6</td>
</tr>
</table>

<p>The allowed bit depths and sample depths for each PNG image
type are listed in 11.2.2: <a href="#11IHDR"><span class=
"xref"><span class="chunk">IHDR</span> Image
header</span></a>.</p>

<p>Greyscale samples represent luminance if the transfer curve is
indicated (by <a href="#11gAMA"><span class=
"chunk">gAMA</span></a>, <a href="#11sRGB"><span class=
"chunk">sRGB</span></a>, or <a href="#11iCCP"><span class=
"chunk">iCCP</span></a>) or device-dependent greyscale if not.
RGB samples represent calibrated colour information if the colour
space is indicated (by <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> and <a href="#11cHRM"><span class=
"chunk">cHRM</span></a>, or <a href="#11sRGB"><span class=
"chunk">sRGB</span></a>, or <a href="#11iCCP"><span class=
"chunk">iCCP</span></a>) or uncalibrated device-dependent colour
if not.</p>

<p>Sample values are not necessarily proportional to light
intensity; the <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> chunk specifies the relationship between
sample values and display output intensity. Viewers are strongly
encouraged to compensate properly. See 4.2: <a href=
"#4Concepts.ColourSpaces"><span class="xref">Colour
spaces</span></a>, 13.13: <a href=
"#13Decoder-gamma-handling"><span class="xref">Decoder gamma
handling</span></a> and Annex C: <a href="#C-GammaAppendix"><span
class="xref">Gamma and chromaticity</span></a>.</p>

<h2><a name="6AlphaRepresentation">6.2 Alpha
representation</a></h2>

<p>In a PNG datastream transparency may be represented in one of
four ways, depending on the PNG image type (see 4.3.2: <a href=
"#4Concepts.Implied-alpha"><span class="xref">Alpha
separation</span></a> and 4.3.5: <a href=
"#4Concepts.Alpha-indexing"><span class="xref">Alpha
compaction</span></a>).</p>

<!-- <ol start="1"> --><ol>
<li>Truecolour with alpha, greyscale with alpha: an alpha channel
is part of the image array.</li>

<li>Truecolour, greyscale: A <a href="#11tRNS"><span class=
"chunk">tRNS</span></a> chunk contains a single pixel value
distinguishing the fully transparent pixels from the fully opaque
pixels.</li>

<li>Indexed-colour: A <a href="#11tRNS"><span class=
"chunk">tRNS</span></a> chunk contains the alpha table that
associates an alpha sample with each palette entry.</li>

<li>Truecolour, greyscale, indexed-colour: there is no <a href=
"#11tRNS"><span class="chunk">tRNS</span></a> chunk present and
all pixels are fully opaque.</li>
</ol>

<p>An alpha channel included in the image array has 8-bit or
16-bit samples, the same size as the other samples. The alpha
sample for each pixel is stored immediately following the
greyscale or RGB samples of the pixel. An alpha value of zero
represents full transparency, and a value of
2<sup>sampledepth</sup> - 1 represents full opacity. Intermediate
values indicate partially transparent pixels that can be
composited against a background image to yield the delivered
image.</p>

<p>The colour values in a pixel are not premultiplied by the
alpha value assigned to the pixel. This rule is sometimes called
"unassociated" or "non-premultiplied" alpha. (Another common
technique is to store sample values premultiplied by the alpha
value; in effect, such an image is already composited against a
black background. PNG does <strong>not</strong> use premultiplied alpha.
In consequence an image editor can take a PNG image and easily
change its transparency.) See 12.4: <a href=
"#12Alpha-channel-creation"><span class="xref">Alpha channel
creation</span></a> and 13.16: <a href=
"#13Alpha-channel-processing"><span class="xref">Alpha channel
processing</span></a>.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="7Transformation">7 Encoding the PNG image as a PNG
datastream</a></h1>

<h2><a name="7Integers-and-byte-order">7.1 Integers and byte
order</a></h2>

<p>All integers that require more than one byte shall be in
network byte order (as illustrated in <a href="#figure71"><span
class="figref">figure 7.1</span></a>): the most significant byte
comes first, then the less significant bytes in descending order
of significance (MSB LSB for two-byte integers, MSB B2 B1 LSB for
four-byte integers). The highest bit (value 128) of a byte is
numbered bit 7; the lowest bit (value 1) is numbered bit 0.
Values are unsigned unless otherwise noted. Values explicitly
noted as signed are represented in two's complement notation.</p>

<p>PNG four-byte unsigned integers are limited to the range 0 to
2<sup>31</sup>-1 to accommodate languages that have difficulty
with unsigned four-byte values. Similarly PNG four-byte signed
integers are limited to the range -(2<sup>31</sup>-1) to
2<sup>31</sup>-1 to accommodate languages that have difficulty
with the value -2<sup>31</sup>.</p>

<p>
<object height="310" width="810" data="figures/fig71.svg" type="image/svg+xml">
  <img height="310" width="810" src="png-figures/fig71.png" alt="Figure 7.1: Integer representation in PNG" /> 
</object>
</p>

<p class="Figuretitle"><a name="figure71">Figure 7.1</a> &mdash;
Integer representation in PNG</p>

<h2><a name="7Scanline">7.2 Scanlines</a></h2>

<p>A PNG image (or pass, see clause&#160;8: <a href=
"#8Interlace"><span class="xref">Interlacing and pass
extraction</span></a>) is a rectangular pixel array, with pixels
appearing left-to-right within each scanline, and scanlines
appearing top-to-bottom. The size of each pixel is determined by
the number of bits per pixel.</p>

<p>Pixels within a scanline are always packed into a sequence of
bytes with no wasted bits between pixels. Scanlines always begin
on byte boundaries. Permitted bit depths and colour types are
restricted so that in all cases the packing is simple and
efficient.</p>

<p>
In PNG images of colour type 0 (greyscale) each pixel is a single sample, which may have precision less than a byte (1, 2, or 4 bits). These samples are packed into bytes with the leftmost sample in the high-order bits of a byte followed by the other samples for the scanline.
</p>
<p>
In PNG images of colour type 3 (indexed-colour) each pixel is a single palette index. These indices are packed into bytes in the same way as the samples for colour type 0.</p>
<p>When there are multiple pixels per byte, some low-order bits
of the last byte of a scanline may go unused. The contents of
these unused bits are not specified.</p>

<p>PNG images that are not indexed-colour images may have sample
values with a bit depth of 16. Such sample values are in network
byte order (MSB first, LSB second). PNG permits multi-sample
pixels only with 8 and 16-bit samples, so multiple samples of a
single pixel are never packed into one byte.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="7Filtering">7.3 Filtering</a></h2>

<p>PNG allows the scanline data to be <strong>filtered</strong> before it
is compressed. Filtering can improve the compressibility of the
data. The filter step itself results in a sequence of bytes of
the same size as the incoming sequence, but in a different
representation, preceded by a filter type byte. Filtering does
not reduce the size of the actual scanline data. All PNG filters
are strictly lossless.</p>

<p>Different filter types can be used for different scanlines,
and the filter algorithm is specified for each scanline by a
filter type byte. The filter type byte is not considered part of
the image data, but it is included in the datastream sent to the
compression step. An intelligent encoder can switch filters from
one scanline to the next. The method for choosing which filter to
employ is left to the encoder.</p>

<p>See clause&#160;9: <a href="#9Filters"><span class=
"xref">Filtering</span></a>.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="8Interlace">8 Interlacing and pass
extraction</a></h1>

<h2><a name="8InterlaceIntro">8.1 Introduction</a></h2>

<p>Pass extraction (see <a href="#figure48"><span class=
"figref">figure 4.8</span></a>) splits a PNG image into a
sequence of reduced images (the interlaced PNG image) where the
first image defines a coarse view and subsequent images enhance
this coarse view until the last image completes the PNG image.
This allows progressive display of the interlaced PNG image by
the decoder and allows images to "fade in" when they are being
displayed on-the-fly. On average, interlacing slightly expands
the datastream size, but it can give the user a meaningful
display much more rapidly.</p>

<h2><a name="8InterlaceMethods">8.2 Interlace methods</a></h2>

<p>Two interlace methods are defined in this International
Standard, methods 0 and 1. Other values of interlace method are
reserved for future standardization (see 4.9: <a href=
"#4Concepts.Registration"><span class="xref">Extension and
registration</span></a>).</p>

<p>With interlace method 0, the null method, pixels are extracted
sequentially from left to right, and scanlines sequentially from
top to bottom. The interlaced PNG image is a single reduced
image.</p>

<p>Interlace method 1, known as Adam7, defines seven distinct
passes over the image. Each pass transmits a subset of the pixels
in the reference image. The pass in which each pixel is
transmitted (numbered from 1 to 7) is defined by replicating the
following 8-by-8 pattern over the entire image, starting at the
upper left corner:</p>

<pre>
   1 6 4 6 2 6 4 6
   7 7 7 7 7 7 7 7
   5 6 5 6 5 6 5 6
   7 7 7 7 7 7 7 7
   3 6 4 6 3 6 4 6
   7 7 7 7 7 7 7 7
   5 6 5 6 5 6 5 6
   7 7 7 7 7 7 7 7
</pre>

<p><a href="#figure48"><span class="figref">Figure 4.8</span></a>
shows the seven passes of interlace method 1. Within each pass,
the selected pixels are transmitted left to right within a
scanline, and selected scanlines sequentially from top to bottom.
For example, pass 2 contains pixels 4, 12, 20, etc. of scanlines
0, 8, 16, etc. (where scanline 0, pixel 0 is the upper left
corner). The last pass contains all of scanlines 1, 3, 5, etc.
The transmission order is defined so that all the scanlines
transmitted in a pass will have the same number of pixels; this
is necessary for proper application of some of the filters. The
interlaced PNG image consists of a sequence of seven reduced
images. For example, if the PNG image is 16 by 16 pixels, then
the third pass will be a reduced image of two scanlines, each
containing four pixels (see <a href="#figure48"><span class=
"figref">figure 4.8</span></a>).</p>

<p>Scanlines that do not completely fill an integral number of
bytes are padded as defined in 7.2: <a href="#7Scanline"><span
class="xref">Scanlines</span></a>.</p>

<p class="Note">NOTE If the reference image contains fewer than
five columns or fewer than five rows, some passes will be
empty.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="9Filters">9 Filtering</a></h1>

<h2><a name="9FtIntro">9.1 Filter methods and filter
types</a></h2>

<p>Filtering transforms the PNG image with the goal of
improving compression. PNG allows for a number of filter methods.
All the reduced
images in an interlaced image shall use a single filter method.
Only filter method 0
is defined by this International Standard. Other filter methods
are reserved for future standardization (see 4.9 <a href=
"#4Concepts.Registration"><span class="xref">Extension and
registration</span></a>).
Filter method 0 provides a set of five filter types,
and individual scanlines in each reduced image may use
different filter types.</p>

<p>PNG imposes no additional restriction on which filter types
can be applied to an interlaced PNG image. However, the filter
types are not equally effective on all types of data. See 12.8:
<a href="#12Filter-selection"><span class="xref">Filter
selection</span></a>.</p>

<p>Filtering transforms the byte sequence in a scanline to an
equal length sequence of bytes preceded by the filter type.
Filter type bytes are associated only with non-empty scanlines.
No filter type bytes are present in an empty pass. See 13.8: <a
href="#13Progressive-display"><span class="xref">Interlacing and
progressive display</span></a>.</p>

<h2><a name="9Filter-types">9.2 Filter types for filter method
0</a></h2>

<p>Filters are applied to <strong>bytes</strong>, not to pixels,
regardless of the bit depth or colour type of the image. The
filters operate on the byte sequence formed by a scanline that
has been represented as described in 7.2: <a href=
"#7Scanline"><span class="xref">Scanlines</span></a>. If the image
includes an alpha channel, the alpha data is filtered in the same
way as the image data.</p>

<p>Filters may use the original values of the following bytes to
generate the new byte value:</p>

<table class="Regular" summary=
"This table defines the variables usedin table 9.1">
<tr>
<td class="Regular"><tt>x</tt> </td>
<td class="Regular">the byte being filtered;</td>
</tr>

<tr>
<td class="Regular"><tt>a</tt> </td>
<td class="Regular">the byte corresponding to x in the pixel immediately before the pixel containing x (or the byte immediately before x, when the bit depth is less than 8);</td>
</tr>

<tr>
<td class="Regular"><tt>b</tt> </td>
<td class="Regular">the byte corresponding to x in the previous scanline;</td>
</tr>

<tr>
<td class="Regular"><tt>c</tt> </td>
<td class="Regular">the byte corresponding to b in the pixel immediately before the pixel containing b (or the byte immediately before b, when the bit depth is less than 8).</td>
</tr>
</table>

<p><a href="#9-figure91"><span class="figref">Figure
9.1</span></a> shows the relative positions of the bytes <tt>x</tt>,
<tt>a</tt>, <tt>b</tt>,
and <tt>c</tt>.</p>

<p>PNG filter method 0 defines five basic filter types as listed
in <a href="#9-table91"><span class="tabref">Table
9.1</span></a>. <tt>Orig(y)</tt> denotes the orginal (unfiltered)
value of byte <tt>y</tt>. <tt>Filt(y)</tt> denotes the value
after a filter has been applied. <tt>Recon(y)</tt> denotes the
value after the corresponding reconstruction function has been
applied. The filter function for the Paeth type
<tt>PaethPredictor</tt> is defined below.</p>

<p>Filter method 0 specifies exactly this set of five filter
types and this shall not be extended.
This ensures that decoders need not decompress the data
to determine whether it contains unsupported filter types:
it is sufficient to check the filter method in <a href="#11IHDR"><span class=
"chunk">IHDR</span></a>.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<table class="Regular" summary=
"This table lists the filter types">
<caption><a name="9-table91"><b>Table 9.1 &mdash; Filter
types</b></a></caption>

<tr>
<th>Type</th>
<th>Name</th>
<th>Filter Function</th>
<th>Reconstruction Function</th>
</tr>

<tr>
<td class="Regular" align="center">0</td>
<td class="Regular">None</td>
<td class="Regular"><tt>Filt(x) = Orig(x)</tt> </td>
<td class="Regular"><tt>Recon(x) = Filt(x)</tt> </td>
</tr>

<tr>
<td class="Regular" align="center">1</td>
<td class="Regular">Sub</td>
<td class="Regular"><tt>Filt(x) = Orig(x) - Orig(a)</tt> </td>
<td class="Regular"><tt>Recon(x) = Filt(x) + Recon(a)</tt> </td>
</tr>

<tr>
<td class="Regular" align="center">2</td>
<td class="Regular">Up</td>
<td class="Regular"><tt>Filt(x) = Orig(x) - Orig(b)</tt> </td>
<td class="Regular"><tt>Recon(x) = Filt(x) + Recon(b)</tt> </td>
</tr>

<tr>
<td class="Regular" align="center">3</td>
<td class="Regular">Average</td>
<td class="Regular"><tt>Filt(x) = Orig(x) - floor((Orig(a) + Orig(b)) /
2)</tt> </td>
<td class="Regular"><tt>Recon(x) = Filt(x) + floor((Recon(a) + Recon(b)) /
2)</tt> </td>
</tr>

<tr>
<td class="Regular" align="center">4</td>
<td class="Regular">Paeth</td>
<td class="Regular"><tt>Filt(x) = Orig(x) - PaethPredictor(Orig(a),
Orig(b), Orig(c))</tt> </td>
<td class="Regular"><tt>Recon(x) = Filt(x) + PaethPredictor(Recon(a), Recon(b),
Recon(c))</tt> </td>
</tr>
</table>

<p>For all filters, the bytes "to the left of" the first pixel in
a scanline shall be treated as being zero. For filters that refer
to the prior scanline, the entire prior scanline and bytes "to
the left of" the first pixel in the prior scanline shall be
treated as being zeroes for the first scanline of a reduced
image.</p>

<p>To reverse the effect of a filter requires the decoded values
of the prior pixel on the same scanline, the pixel immediately
above the current pixel on the prior scanline, and the pixel just
to the left of the pixel above.</p>

<p>Unsigned arithmetic modulo 256 is used, so that both the
inputs and outputs fit into bytes. Filters are applied to each
byte regardless of bit depth. The sequence of <tt>Filt</tt>
values is transmitted as the filtered scanline.</p>

<h2><a name="9Filter-type-3-Average">9.3 Filter type 3:
Average</a></h2>

<p>The sum <tt>Orig(a) + Orig(b)</tt> shall be performed without
overflow (using at least nine-bit arithmetic). <tt>floor()</tt>
indicates that the result of the division is rounded to the next
lower integer if fractional; in other words, it is an integer
division or right shift operation.</p>

<h2><a name="9Filter-type-4-Paeth">9.4 Filter type 4:
Paeth</a></h2>

<p>The Paeth filter function computes a simple linear function of
the three neighbouring pixels (left, above, upper left), then
chooses as predictor the neighbouring pixel closest to the
computed value. The algorithm used in this International Standard
is an adaptation of the technique due to Alan W. Paeth <a href=
"#G-PAETH"><span class="bibref">[PAETH]</span></a>.</p>

<p>The PaethPredictor function is defined in the code below. The
logic of the function and the locations of the bytes <tt>a</tt>,
<tt>b</tt>, <tt>c</tt>, and <tt>x</tt> are shown in <a href=
"#9-figure91"><span class="figref">figure 9.1</span></a>.
<tt>Pr</tt> is the predictor for byte <tt>x</tt>.</p>

<pre>
    p = a + b - c
    pa = abs(p - a)
    pb = abs(p - b)
    pc = abs(p - c)
    if pa &lt;= pb and pa &lt;= pc then Pr = a
    else if pb &lt;= pc then Pr = b
    else Pr = c
    return Pr
</pre>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p><a name="9-figure91">
<object height="360" width="640" data="figures/fig91.svg" type="image/svg+xml">
  <img height="360" width="640" src="png-figures/fig91.png" alt="Figure 9.1: The PaethPredictor
function" />
</object>
</a></p>

<p class="Figuretitle"><b>Figure 9.1: The PaethPredictor
function</b></p>

<p>The calculations within the PaethPredictor function shall be
performed exactly, without overflow.</p>

<p><strong>The order in which the comparisons are performed is
critical and shall not be altered.</strong> The function tries to
establish in which of the three directions (vertical, horizontal,
or diagonal) the gradient of the image is smallest.</p>

<p>Exactly the same PaethPredictor function is used by both
encoder and decoder.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="10Compression">10 Compression</a></h1>

<h2><a name="10CompressionCM0">10.1 Compression method 0</a></h2>

<p>Only PNG compression method 0 is defined by this International
Standard. Other values of compression method are reserved for
future standardization (see 4.9: <a href=
"#4Concepts.Registration"><span class="xref">Extension and
registration</span></a>). PNG compression method 0 is
deflate/inflate compression with a sliding window
(which is an upper bound on the distances appearing in the
deflate stream) of at most
32768 bytes. Deflate compression is an LZ77 derivative <a href=
"#G-ZL"><span class="bibref">[ZL]</span></a>.</p>

<p>Deflate-compressed datastreams within PNG are stored in the
"zlib" format, which has the structure:</p>

<table class="Regular"  summary=
"This table gives the structure of the zlib format">
<tr>
<td class="Regular">zlib compression method/flags code</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Additional flags/check bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Compressed data blocks</td>
<td class="Regular">n bytes</td>
</tr>

<tr>
<td class="Regular">Check value</td>
<td class="Regular">4 bytes</td>
</tr>
</table>

<p>Further details on this format are given in the zlib
specification <a href="#2-RFC-1950"><span class=
"NormRef">[RFC-1950]</span></a>.</p>

<p>For PNG compression method 0, the zlib compression
method/flags code shall specify method code 8 (deflate
compression) and an LZ77 window size of not more than 32768
bytes. The zlib compression method number is not the same as the
PNG compression method number in the <a href=
"#11IHDR"><span class="chunk">IHDR</span></a> chunk (see 11.2.2
<a href="#11IHDR"><span class="xref"><span class=
"chunk">IHDR</span> Image header</span></a>). The additional
flags shall not specify a preset dictionary.</p>

<p>If the data to be compressed contain 16384 bytes or fewer, the
PNG encoder may set the window size by rounding up to a power of
2 (256 minimum). This decreases the memory required for both
encoding and decoding, without adversely affecting the
compression ratio.</p>

<p>The compressed data within the zlib datastream are stored as a
series of blocks, each of which can represent raw (uncompressed)
data, LZ77-compressed data encoded with fixed Huffman codes, or
LZ77-compressed data encoded with custom Huffman codes. A marker
bit in the final block identifies it as the last block, allowing
the decoder to recognize the end of the compressed datastream.
Further details on the compression algorithm and the encoding are
given in the deflate specification <a href="#2-RFC-1951"><span
class="NormRef">[RFC-1951]</span></a>.</p>

<p>The check value stored at the end of the zlib datastream is
calculated on the uncompressed data represented by the
datastream. The algorithm used to calculate this is not the same
as the CRC calculation used for PNG chunk CRC field values. The
zlib check value is useful mainly as a cross-check that the
deflate and inflate algorithms are implemented correctly.
Verifying the individual PNG chunk CRCs provides confidence that
the PNG datastream has been transmitted undamaged.</p>

<h2><a name="10CompressionFSL">10.2 Compression of the sequence
of filtered scanlines</a></h2>

<p>The sequence of filtered scanlines is compressed and the
resulting data stream is split into <a href="#11IDAT"><span
class="chunk">IDAT</span></a> chunks. The concatenation of the
contents of all the <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunks makes up a zlib datastream. This
datastream decompresses to filtered image data.</p>

<p>It is important to emphasize that the boundaries between <a
href="#11IDAT"><span class="chunk">IDAT</span></a> chunks are
arbitrary and can fall anywhere in the zlib datastream. There is
not necessarily any correlation between <a href="#11IDAT"><span
class="chunk">IDAT</span></a> chunk boundaries and deflate block
boundaries or any other feature of the zlib data. For example, it
is entirely possible for the terminating zlib check value to be
split across <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunks.</p>

<p>Similarly, there is no required correlation between the
structure of the image data (i.e., scanline boundaries) and
deflate block boundaries or <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunk boundaries. The complete filtered
PNG image is represented by a single zlib datastream that is
stored in a number of <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunks.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="10CompressionOtherUses">10.3 Other uses of
compression</a></h2>

<p>PNG also uses compression method 0 in <a href="#11iTXt"><span
class="chunk">iTXt</span></a>, <a href="#11iCCP"><span class=
"chunk">iCCP</span></a>, and <a href="#11zTXt"><span class=
"chunk">zTXt</span></a> chunks. Unlike the image data, such
datastreams are not split across chunks; each such chunk contains
an independent zlib datastream (see 10.1: <a href=
"#10CompressionCM0"><span class="xref">Compression method
0</span></a>).</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="11Chunks">11 Chunk specifications</a></h1>

<h2><a name="11Introduction">11.1 Introduction</a></h2>

<p>The PNG datastream consists of a PNG signature (see 5.2: <a
href="#5PNG-file-signature"><span class="xref">PNG
signature</span></a>) followed by a sequence of chunks. Each
chunk has a chunk type which specifies its function. This clause
defines the PNG chunk types standardized in this International
Standard. The PNG datastream structure is defined in clause&#160;5: <a
href="#5DataRep"><span class="xref">Datastream
structure</span></a>. This also defines the order in which chunks
may appear. For details specific to encoders see 12.11: <a href=
"#12Chunk-processing"><span class="xref">Chunking</span></a>.
For details specific to decoders see 13.5: <a href=
"#13Chunking"><span class="xref">Chunking</span></a>.</p>

<h2><a name="11Critical-chunks">11.2 Critical chunks</a></h2>

<h3><a name="11CcGen">11.2.1 General</a></h3>

<p>Critical chunks are those chunks that are absolutely required
in order to successfully decode a PNG image from a PNG
datastream. Extension chunks may be defined as critical chunks
(see clause&#160;14: <a href="#14EditorsExt"><span class=
"xref">Editors and extensions</span></a>), though this practice
is strongly discouraged.</p>

<p>A valid PNG datastream shall begin with a PNG signature,
immediately followed by an <a href="#11IHDR"><span class=
"chunk">IHDR</span></a> chunk, then one or more <a href=
"#11IDAT"><span class="chunk">IDAT</span></a> chunks, and shall
end with an <a href="#11IEND"><span class="chunk">IEND</span></a>
chunk. Only one <a href="#11IHDR"><span class=
"chunk">IHDR</span></a> chunk and one <a href="#11IEND"><span
class="chunk">IEND</span></a> chunk are allowed in a PNG
datastream.</p>

<h3><a name="11IHDR">11.2.2 <span class="chunk">IHDR</span> Image
header</a></h3>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
73 72 68 82
</pre>

<p>The <span class="chunk">IHDR</span> chunk shall be the first
chunk in the PNG datastream. It contains:</p>

<table class="Regular"  summary=
"This table defines the IHDR chunk">
<tr>
<td class="Regular">Width</td>
<td class="Regular">4 bytes</td>
</tr>

<tr>
<td class="Regular">Height</td>
<td class="Regular">4 bytes</td>
</tr>

<tr>
<td class="Regular">Bit depth</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Colour type</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Compression method</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Filter method</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Interlace method</td>
<td class="Regular">1 byte</td>
</tr>
</table>

<p>Width and height give the image dimensions in pixels. They are
PNG four-byte unsigned integers. Zero is an invalid
value.</p>

<p>Bit depth is a single-byte integer giving the number of bits
per sample or per palette index (not per pixel). Valid values are
1, 2, 4, 8, and 16, although not all values are allowed for all
colour types. See 6.1: <a href="#6Colour-values"><span class=
"xref">Colour types and values</span></a>.</p>

<p>Colour type is a single-byte integer that defines the PNG
image type. Valid values are 0, 2, 3, 4, and 6.</p>

<p>Bit depth restrictions for each colour type are imposed to
simplify implementations and to prohibit combinations that do not
compress well. The allowed combinations are defined in <a href=
"#table111"><span class="tabref">Table 11.1</span></a>.</p>

<table class="Regular" summary=
"This table defines the colour types">
<caption><a name="table111"><b>Table 11.1 &mdash; Allowed
combinations of colour type and bit depth</b></a></caption>

<tr>
<th>PNG image type</th>
<th>Colour type</th>
<th>Allowed bit depths</th>
<th>Interpretation</th>
</tr>

<tr>
<td class="Regular">Greyscale</td>
<td class="Regular" align="center">0</td>
<td class="Regular">1, 2, 4, 8, 16</td>
<td class="Regular">Each pixel is a greyscale sample</td>
</tr>

<tr>
<td class="Regular">Truecolour</td>
<td class="Regular" align="center">2</td>
<td class="Regular">8, 16</td>
<td class="Regular">Each pixel is an R,G,B triple</td>
</tr>

<tr>
<td class="Regular">Indexed-colour</td>
<td class="Regular" align="center">3</td>
<td class="Regular">1, 2, 4, 8</td>
<td class="Regular">Each pixel is a palette index; a <a href="#11PLTE"><span
class="chunk">PLTE</span></a> chunk shall appear.</td>
</tr>

<tr>
<td class="Regular">Greyscale with alpha</td>
<td class="Regular" align="center">4</td>
<td class="Regular">8, 16</td>
<td class="Regular">Each pixel is a greyscale sample followed by an alpha
sample.</td>
</tr>

<tr>
<td class="Regular">Truecolour with alpha</td>
<td class="Regular" align="center">6</td>
<td class="Regular">8, 16</td>
<td class="Regular">Each pixel is an R,G,B triple followed by an alpha
sample.</td>
</tr>
</table>

<p>The sample depth is the same as the bit depth except in the
case of indexed-colour PNG images (colour type 3), in which the
sample depth is always 8 bits (see 4.4: <a href=
"#4Concepts.PNGImage"><span class="xref">PNG image</span></a>).</p>

<p>Compression method is a single-byte integer that indicates the
method used to compress the image data. Only compression method 0
(deflate/inflate compression with a sliding window of at most
32768 bytes) is defined in this International Standard. All
conforming PNG images shall be compressed with this scheme.</p>

<p>Filter method is a single-byte integer that indicates the
preprocessing method applied to the image data before
compression. Only filter method 0 (adaptive filtering with five
basic filter types) is defined in this International Standard.
See clause&#160;9: <a href="#9Filters"><span class=
"xref">Filtering</span></a> for details.</p>

<p>Interlace method is a single-byte integer that indicates the
transmission order of the image data. Two values are defined in
this International Standard: 0 (no interlace) or 1 (Adam7
interlace). See clause&#160;8: <a href="#8Interlace"><span class=
"xref">Interlacing and pass extraction</span></a> for
details.</p>

<h3><a name="11PLTE">11.2.3 <span class="chunk">PLTE</span>
Palette</a></h3>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
80 76 84 69
</pre>

<p>The <span class="chunk">PLTE</span> chunk contains from 1 to
256 palette entries, each a three-byte series of the form:</p>

<table class="Regular" summary=
"This table defines the PLTE palette table entries">
<tr>
<td class="Regular">Red</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Green</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Blue</td>
<td class="Regular">1 byte</td>
</tr>
</table>

<p>The number of entries is determined from the chunk length. A
chunk length not divisible by 3 is an error.</p>

<p>This chunk shall appear for colour type 3, and may appear for
colour types 2 and 6; it shall not appear for colour types 0 and
4. There shall not be more than one <span class=
"chunk">PLTE</span> chunk.</p>

<p>For colour type 3 (indexed-colour), the <span class=
"chunk">PLTE</span> chunk is required. The first entry in <span
class="chunk">PLTE</span> is referenced by pixel value 0, the
second by pixel value 1, etc. The number of palette entries shall
not exceed the range that can be represented in the image bit
depth (for example, 2<sup>4</sup> = 16 for a bit depth of 4). It
is permissible to have fewer entries than the bit depth would
allow. In that case, any out-of-range pixel value found in the
image data is an error.</p>

<p>For colour types 2 and 6 (truecolour and truecolour with
alpha), the <span class="chunk">PLTE</span> chunk is optional. If
present, it provides a suggested set of colours (from 1 to 256)
to which the truecolour image can be quantized if it cannot be
displayed directly. It is, however, recommended that the <a href=
"#11sPLT"><span class="chunk">sPLT</span></a> chunk be used for
this purpose, rather than the <span class="chunk">PLTE</span>
chunk. If neither <span class="chunk">PLTE</span> nor <a href=
"#11sPLT"><span class="chunk">sPLT</span></a> chunks are present
and the image cannot be displayed directly, quantization has to
be done by the viewing system. However, it is often preferable
for the selection of colours to be done once by the PNG encoder.
(See 12.6: <a href="#12Suggested-palettes"><span class=
"xref">Suggested palettes</span></a>.)</p>

<p>Note that the palette uses 8 bits (1 byte) per sample
regardless of the image bit depth. In particular,
the palette is 8 bits deep even when it is a suggested
quantization of a 16-bit truecolour image.</p>

<p>There is no requirement that the palette entries all be used
by the image, nor that they all be different.</p>

<h3><a name="11IDAT">11.2.4 <span class="chunk">IDAT</span> Image
data</a></h3>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
73 68 65 84
</pre>

<p>The <span class="chunk">IDAT</span> chunk contains the actual
image data which is the output stream of the compression
algorithm. See clause&#160;9: <a href="#9Filters"><span class=
"xref">Filtering</span></a> and clause&#160;10: <a href=
"#10Compression"><span class="xref">Compression</span></a> for
details.</p>

<p>There may be multiple <span class="chunk">IDAT</span> chunks;
if so, they shall appear consecutively with no other intervening
chunks. The compressed datastream is then the concatenation of
the contents of the data fields of all the <span class=
"chunk">IDAT</span> chunks.</p>

<h3><a name="11IEND">11.2.5 <span class="chunk">IEND</span> Image
trailer</a></h3>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
73 69 78 68
</pre>

<p>The <span class="chunk">IEND</span> chunk marks the end of the
PNG datastream. The chunk's data field is empty.</p>

<h2><a name="11Ancillary-chunks">11.3 Ancillary chunks</a></h2>

<h3><a name="11AcGen">11.3.1 General</a></h3>

<p>The ancillary chunks defined in this International Standard
are listed in the order in 4.7.2: <a href=
"#4Concepts.FormatTypes"><span class="xref">Chunk
types</span></a>. This is not the order in which they appear in a
PNG datastream. Ancillary chunks may be ignored by a decoder. For
each ancillary chunk, the actions described are under the
assumption that the decoder is not ignoring the chunk.</p>

<h3><a name="11transinfo">11.3.2 Transparency
information</a></h3>

<h4><a name="11tRNS">11.3.2.1 <span class="chunk">tRNS</span>
Transparency</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
116 82 78 83
</pre>

<p>The <span class="chunk">tRNS</span> chunk specifies either
alpha values that are associated with palette entries (for
indexed-colour images) or a single transparent colour (for
greyscale and truecolour images). The <span class=
"chunk">tRNS</span> chunk contains: 
<!-- ************Page Break******************* -->
</p>

<!-- ************Page Break******************* -->
<table class="Regular" summary=
"This table defines the tRNS chunk">
<tr>
<th colspan="2">Colour type 0</th>
</tr>

<tr>
<td class="Regular">Grey sample value</td>
<td class="Regular">2 bytes</td>
</tr>

<tr>
<th colspan="2">Colour type 2</th>
</tr>

<tr>
<td class="Regular">Red sample value</td>
<td class="Regular">2 bytes</td>
</tr>

<tr>
<td class="Regular">Blue sample value</td>
<td class="Regular">2 bytes</td>
</tr>

<tr>
<td class="Regular">Green sample value</td>
<td class="Regular">2 bytes</td>
</tr>

<tr>
<th colspan="2">Colour type 3</th>
</tr>

<tr>
<td class="Regular">Alpha for palette index 0</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Alpha for palette index 1</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">...etc...</td>
<td class="Regular">1 byte</td>
</tr>
</table>

<p>For colour type 3 (indexed-colour), the <span class=
"chunk">tRNS</span> chunk contains a series of one-byte alpha
values, corresponding to entries in the <a href="#11PLTE"><span
class="chunk">PLTE</span></a> chunk. Each entry indicates that
pixels of the corresponding palette index shall be treated as
having the specified alpha value. Alpha values have the same
interpretation as in an 8-bit full alpha channel: 0 is fully
transparent, 255 is fully opaque, regardless of image bit depth.
The <span class="chunk">tRNS</span> chunk shall not contain more
alpha values than there are palette entries, but a <span class=
"chunk">tRNS</span> chunk may contain fewer values than there are
palette entries. In this case, the alpha value for all remaining
palette entries is assumed to be 255. In the common case in which
only palette index 0 need be made transparent, only a one-byte
<span class="chunk">tRNS</span> chunk is needed, and when all
palette indices are opaque, the <span class="chunk">tRNS</span>
chunk may be omitted.</p>

<p>For colour types 0 or 2, two bytes per sample are used
regardless of the image bit depth (see 7.1: <a href=
"#7Integers-and-byte-order"><span class="xref">Integers and byte
order</span></a>). Pixels of the specified grey sample value or
RGB sample values are treated as transparent (equivalent to alpha
value 0); all other pixels are to be treated as fully opaque
(alpha value 2<sup>bitdepth</sup>-1). If the image bit depth is
less than 16, the least significant bits are used and the others
are 0.</p>

<p>A <span class="chunk">tRNS</span> chunk shall not appear for
colour types 4 and 6, since a full alpha channel is already
present in those cases.</p>

<p class="Note">NOTE For 16-bit greyscale or truecolour data,
only pixels matching the entire 16-bit values in <span class=
"chunk">tRNS</span> chunks are transparent. Decoders have to
postpone any sample depth rescaling until after the pixels have
been tested for transparency.</p>

<h3><a name="11addnlcolinfo">11.3.3 Colour space
information</a></h3>

<h4><a name="11cHRM">11.3.3.1 <span class="chunk">cHRM</span>
Primary chromaticities and white point</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
99 72 82 77
</pre>

<p>The <span class="chunk">cHRM</span> chunk may be used to
specify the 1931 CIE <i>x,y</i> chromaticities of the red,
green, and blue display primaries used in the image, and the referenced
white point. See Annex C: <a href="#C-GammaAppendix"><span class=
"xref">Gamma and chromaticity</span></a> for more information.
The <a href="#11iCCP"><span class="chunk">iCCP</span></a> and <a
href="#11sRGB"><span class="chunk">sRGB</span></a> chunks provide
more sophisticated support for colour management and control.</p>

<p>The <span class="chunk">cHRM</span> chunk contains:</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<table class="Regular" summary=
"This table defines the cHRM chunk">
<tr>
<td class="Regular">White point x</td>
<td class="Regular">4 bytes</td>
</tr>

<tr>
<td class="Regular">White point y</td>
<td class="Regular">4 bytes</td>
</tr>

<tr>
<td class="Regular">Red x</td>
<td class="Regular">4 bytes</td>
</tr>

<tr>
<td class="Regular">Red y</td>
<td class="Regular">4 bytes</td>
</tr>

<tr>
<td class="Regular">Green x</td>
<td class="Regular">4 bytes</td>
</tr>

<tr>
<td class="Regular">Green y</td>
<td class="Regular">4 bytes</td>
</tr>

<tr>
<td class="Regular">Blue x</td>
<td class="Regular">4 bytes</td>
</tr>

<tr>
<td class="Regular">Blue y</td>
<td class="Regular">4 bytes</td>
</tr>
</table>

<p>Each value is encoded as a four-byte PNG unsigned integer,
representing the <i>x</i> or <i>y</i> value times 100000.</p>

<p>EXAMPLE A value of 0.3127 would be stored as the integer
31270.</p>

<p>The <span class="chunk">cHRM</span> chunk is allowed in all
PNG datastreams, although it is of little value for greyscale
images.</p>

<p>An <a href="#11sRGB"><span class="chunk">sRGB</span></a> chunk
or <a href="#11iCCP"><span class="chunk">iCCP</span></a> chunk,
when present and recognized, overrides the <span class=
"chunk">cHRM</span> chunk.</p>

<h4><a name="11gAMA">11.3.3.2 <span class="chunk">gAMA</span>
Image gamma</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
103 65 77 65
</pre>

<p>The <span class="chunk">gAMA</span> chunk specifies the
relationship between the image samples and the desired display
output intensity. Gamma is defined in 3.1.20: <a href=
"#3gamma">gamma</a>.</p>

<p>In fact specifying the desired display output intensity is
insufficient. It is also necessary to specify the viewing
conditions under which the output is desired. For <span class=
"chunk">gAMA</span> these are the reference viewing conditions of
the sRGB specification <a href="#2-IEC-61966-2-1"><span class=
"NormRef">[IEC 61966-2-1]</span></a>, which are based on ISO 3664
<a href="#G-ISO-3664"><span class="bibref">[ISO-3664]</span></a>.
Adjustment for different viewing conditions is normally handled
by a Colour Management System. If the adjustment is not
performed, the error is usually small. Applications desiring high
colour fidelity may wish to use an <a href="#11sRGB"><span class=
"chunk">sRGB</span></a> chunk or <a href="#11iCCP"><span class=
"chunk">iCCP</span></a> chunk.</p>

<p>The <span class="chunk">gAMA</span> chunk contains:</p>

<table class="Regular" summary=
"This table defines the gAMA chunk">
<tr>
<td class="Regular">Image gamma</td>
<td class="Regular">4 bytes</td>
</tr>
</table>

<p>The value is encoded as a four-byte PNG unsigned integer,
representing gamma times 100000.</p>

<p>EXAMPLE A gamma of 1/2.2 would be stored as the integer
45455.</p>

<p>See 12.2: <a href="#12Encoder-gamma-handling"><span class=
"xref">Encoder gamma handling</span></a> and 13.13: <a href=
"#13Decoder-gamma-handling"><span class="xref">Decoder gamma
handling</span></a> for more information.</p>

<p>An <a href="#11sRGB"><span class="chunk">sRGB</span></a> chunk
or <a href="#11iCCP"><span class="chunk">iCCP</span></a> chunk,
when present and recognized, overrides the <span class=
"chunk">gAMA</span> chunk.</p>

<h4><a name="11iCCP">11.3.3.3 <span class="chunk">iCCP</span>
Embedded ICC profile</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
105 67 67 80
</pre>

<p>The <span class="chunk">iCCP</span> chunk contains:</p>

<table class="Regular" summary=
"This table defines the iCCP chunk">
<tr>
<td class="Regular">Profile name</td>
<td class="Regular">1-79 bytes (character string)</td>
</tr>

<tr>
<td class="Regular">Null separator</td>
<td class="Regular">1 byte (null character)</td>
</tr>

<tr>
<td class="Regular">Compression method</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Compressed profile</td>
<td class="Regular">n bytes</td>
</tr>
</table>

<p>The profile name may be any convenient name for referring to
the profile. It is case-sensitive. Profile names shall contain
only printable Latin-1 characters and spaces (only character
codes 32-126 and 161-255 decimal are allowed). Leading, trailing,
and consecutive spaces are not permitted. The only compression
method defined in this International Standard is method 0 (zlib
datastream with deflate compression, see 10.3: <a href=
'#10CompressionOtherUses'><span class="xref">Other uses of
compression</span></a>). The compression method entry is followed
by a compressed profile that makes up the remainder of the chunk.
Decompression of this datastream yields the embedded ICC
profile.</p>

<p>If the <span class="chunk">iCCP</span> chunk is present, the
image samples conform to the colour space represented by the
embedded ICC profile as defined by the International Color
Consortium <a href="#G-ICC"><span class=
"bibref">[ICC]</span></a>. The colour space of the ICC profile
shall be an RGB colour space for colour images (PNG colour types
2, 3, and 6), or a greyscale colour space for greyscale images
(PNG colour types 0 and 4). A PNG encoder that writes the <span
class="chunk">iCCP</span> chunk is encouraged to also write <a
href="#11gAMA"><span class="chunk">gAMA</span></a> and <a href=
"#11cHRM"><span class="chunk">cHRM</span></a> chunks that
approximate the ICC profile, to provide compatibility with
applications that do not use the <span class="chunk">iCCP</span>
chunk. When the <span class="chunk">iCCP</span> chunk is present,
PNG decoders that recognize it and are capable of colour
management <a href="#G-ICC"><span class="bibref">[ICC]</span></a>
shall ignore the <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> and <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunks and use the <span class=
"chunk">iCCP</span> chunk instead and interpret it according to
<a href="#2-ICC-1"><span class="NormRef">[ICC-1]</span></a> and
<a href="#2-ICC-1A"><span class="NormRef">[ICC-1A]</span></a>.
PNG decoders that are used in an environment that is incapable of
full-fledged colour management should use the <a href=
"#11gAMA"><span class="chunk">gAMA</span></a> and <a href=
"#11cHRM"><span class="chunk">cHRM</span></a> chunks if
present.</p>

<p>A PNG datastream should contain at most one embedded profile,
whether specified explicitly with an <span class=
"chunk">iCCP</span> chunk or implicitly with an <a href=
"#11sRGB"><span class="chunk">sRGB</span></a> chunk.</p>

<h4><a name="11sBIT">11.3.3.4 <span class="chunk">sBIT</span>
Significant bits</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
115 66 73 84
</pre>

<p>To simplify decoders, PNG specifies that only certain sample
depths may be used, and further specifies that sample values
should be scaled to the full range of possible values at the
sample depth. The <a href="#11sBIT"><span class=
"chunk">sBIT</span></a> chunk defines the original number of
significant bits (which can be less than or equal to the sample
depth). This allows PNG decoders to recover the original data
losslessly even if the data had a sample depth not directly
supported by PNG.</p>

<p>The <span class="chunk">sBIT</span> chunk contains:</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<table class="Regular" summary=
"This table defines the sBIT chunk">
<tr>
<th colspan="2">Colour type 0</th>
</tr>

<tr>
<td class="Regular">significant greyscale bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<th colspan="2">Colour types 2 and 3</th>
</tr>

<tr>
<td class="Regular">significant red bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">significant green bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">significant blue bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<th colspan="2">Colour type 4</th>
</tr>

<tr>
<td class="Regular">significant greyscale bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">significant alpha bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<th colspan="2">Colour type 6</th>
</tr>

<tr>
<td class="Regular">significant red bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">significant green bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">significant blue bits</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">significant alpha bits</td>
<td class="Regular">1 byte</td>
</tr>
</table>

<p>Each depth specified in <span class="chunk">sBIT</span> shall
be greater than zero and less than or equal to the sample depth
(which is 8 for indexed-colour images, and the bit depth given in
<a href="#11IHDR"><span class="chunk">IHDR</span></a> for other
colour types).
Note that <span class="chunk">sBIT</span> does not provide a sample depth
for the alpha channel that is implied by a
<a href="#11tRNS"><span class=
"chunk">tRNS</span></a> chunk; in that case, all of the sample bits of
the alpha channel are to be treated as significant. If the <span
class="chunk">sBIT</span> chunk is not present, then all of the
sample bits of all channels are to be treated as significant.</p>

<h4><a name="11sRGB">11.3.3.5 <span class="chunk">sRGB</span>
Standard RGB colour space</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
115 82 71 66
</pre>

<p>If the <span class="chunk">sRGB</span> chunk is present, the
image samples conform to the sRGB colour space <a href=
"#2-IEC-61966-2-1"><span class="NormRef">[IEC
61966-2-1]</span></a> and should be displayed using the specified
rendering intent defined by the International Color Consortium <a
href="#2-ICC-1"><span class="NormRef">[ICC-1]</span></a> and <a
href="#2-ICC-1A"><span class="NormRef">[ICC-1A]</span></a>.</p>

<p>The <span class="chunk">sRGB</span> chunk contains:</p>

<table class="Regular" summary=
"This table defines the sRGB chunk">
<tr>
<td class="Regular">Rendering intent</td>
<td class="Regular">1 byte</td>
</tr>
</table>

<p>The following values are defined for rendering intent:</p>

<table class="Regular" summary=
"This table defines the values of rendering intent in the sRGB chunk">
<tr>
<td class="Regular">0</td>
<td class="Regular">Perceptual</td>
<td class="Regular">for images preferring good adaptation to the output device
gamut at the expense of colorimetric accuracy, such as
photographs.</td>
</tr>

<tr>
<td class="Regular">1</td>
<td class="Regular">Relative colorimetric</td>
<td class="Regular">for images requiring colour appearance matching (relative to
the output device white point), such as logos.</td>
</tr>

<tr>
<td class="Regular">2</td>
<td class="Regular">Saturation</td>
<td class="Regular">for images preferring preservation of saturation at the
expense of hue and lightness, such as charts and graphs.</td>
</tr>

<tr>
<td class="Regular">3</td>
<td class="Regular">Absolute colorimetric</td>
<td class="Regular">for images requiring preservation of absolute colorimetry,
such as previews of images destined for a different output device
(proofs).</td>
</tr>
</table>

<p>It is recommended that a PNG encoder that writes the <span
class="chunk">sRGB</span> chunk also write a <a href=
"#11gAMA"><span class="chunk">gAMA</span></a> chunk (and
optionally a <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunk) for compatibility with decoders
that do not use the <span class="chunk">sRGB</span> chunk. Only
the following values shall be used.</p>

<table class="Regular" summary=
"This table defines the gAMA and cHRM values for sRGB">
<tr>
<th colspan="2"><a href="#11gAMA"><span class=
"chunk">gAMA</span></a> </th>
</tr>

<tr>
<td class="Regular">Gamma</td>
<td class="Regular">45455</td>
</tr>

<tr>
<th colspan="2"><a href="#11cHRM"><span class=
"chunk">cHRM</span></a> </th>
</tr>

<tr>
<td class="Regular">White point x</td>
<td class="Regular">31270</td>
</tr>

<tr>
<td class="Regular">White point y</td>
<td class="Regular">32900</td>
</tr>

<tr>
<td class="Regular">Red x</td>
<td class="Regular">64000</td>
</tr>

<tr>
<td class="Regular">Red y</td>
<td class="Regular">33000</td>
</tr>

<tr>
<td class="Regular">Green x</td>
<td class="Regular">30000</td>
</tr>

<tr>
<td class="Regular">Green y</td>
<td class="Regular">60000</td>
</tr>

<tr>
<td class="Regular">Blue x</td>
<td class="Regular">15000</td>
</tr>

<tr>
<td class="Regular">Blue y</td>
<td class="Regular">6000</td>
</tr>
</table>

<p>When the <span class="chunk">sRGB</span> chunk is present, it
is recommended that decoders that recognize it and are capable of
colour management <a href="#G-ICC"><span class=
"bibref">[ICC]</span></a> ignore the <a href="#11gAMA"><span
class="chunk">gAMA</span></a> and <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunks and use the <span class=
"chunk">sRGB</span> chunk instead. Decoders that recognize the
<span class="chunk">sRGB</span> chunk but are not capable of
colour management <a href="#G-ICC"><span class=
"bibref">[ICC]</span></a> are recommended to ignore the <a href=
"#11gAMA"><span class="chunk">gAMA</span></a> and <a href=
"#11cHRM"><span class="chunk">cHRM</span></a> chunks, and use the
values given above as if they had appeared in <a href=
"#11gAMA"><span class="chunk">gAMA</span></a> and <a href=
"#11cHRM"><span class="chunk">cHRM</span></a> chunks.</p>

<p>It is recommended that the <span class="chunk">sRGB</span> and
<a href="#11iCCP"><span class="chunk">iCCP</span></a> chunks do
not both appear in a PNG datastream.</p>

<h3><a name="11textinfo">11.3.4 Textual information</a></h3>

<h4><a name="11textIntro">11.3.4.1 Introduction</a></h4>

<p>PNG provides the <a href="#11tEXt"><span class=
"chunk">tEXt</span></a>, <a href="#11iTXt"><span class=
"chunk">iTXt</span></a>, and <a href="#11zTXt"><span class=
"chunk">zTXt</span></a> chunks for storing text strings
associated with the image, such as an image description or
copyright notice. Keywords are used to indicate what each text
string represents. Any number of such text chunks may appear, and
more than one with the same keyword is permitted.</p>

<h4><a name="11keywords">11.3.4.2 Keywords and text
strings</a></h4>

<p>The following keywords are predefined and should be used where
appropriate.</p>

<table class="Regular" summary=
"This table defines the keywords defined for tEXt, iTXt and zTXt chunks">
<tr>
<td class="Regular">Title</td>
<td class="Regular">Short (one line) title or caption for image</td>
</tr>

<tr>
<td class="Regular">Author</td>
<td class="Regular">Name of image's creator</td>
</tr>

<tr>
<td class="Regular">Description</td>
<td class="Regular">Description of image (possibly long)</td>
</tr>

<tr>
<td class="Regular">Copyright</td>
<td class="Regular">Copyright notice</td>
</tr>

<tr>
<td class="Regular">Creation Time</td>
<td class="Regular">Time of original image creation</td>
</tr>

<tr>
<td class="Regular">Software</td>
<td class="Regular">Software used to create the image</td>
</tr>

<tr>
<td class="Regular">Disclaimer</td>
<td class="Regular">Legal disclaimer</td>
</tr>

<tr>
<td class="Regular">Warning</td>
<td class="Regular">Warning of nature of content</td>
</tr>

<tr>
<td class="Regular">Source</td>
<td class="Regular">Device used to create the image</td>
</tr>

<tr>
<td class="Regular">Comment</td>
<td class="Regular">Miscellaneous comment</td>
</tr>
</table>

<p>Other keywords may be defined for other purposes. Keywords of
general interest can be registered with the PNG Registration
Authority (see 4.9 <a href="#4Concepts.Registration"><span class=
"xref">Extension and registration</span></a>). It is also
permitted to use private unregistered keywords. (Private keywords
should be reasonably self-explanatory, in order to minimize the
chance that the same keyword is used for incompatible purposes by
different people.)</p>

<p>Keywords shall contain only printable Latin-1 <a href=
"#2-ISO-8859-1"><span class="NormRef">[ISO-8859-1]</span></a>
characters and spaces; that is, only character codes 32-126 and
161-255 decimal are allowed. To reduce the chances for human
misreading of a keyword, leading spaces, trailing spaces,
and consecutive spaces are not permitted in keywords, nor is the
non-breaking space (code 160) since it is visually
indistinguishable from an ordinary space.</p>

<p>Keywords shall be spelled exactly as registered, so that
decoders can use simple literal comparisons when looking for
particular keywords. In particular, keywords are considered
case-sensitive. Keywords are restricted to 1 to 79 bytes in
length.</p>

<p>For the Creation Time keyword, the date format defined in
section&#160;5.2.14 of RFC 1123 is suggested, but not required <a
href="#2-RFC-1123"><span class=
"NormRef">[RFC-1123]</span></a>.</p>

<p>In the <a href="#11tEXt"><span class="chunk">tEXt</span></a>
and <a href="#11zTXt"><span class="chunk">zTXt</span></a> chunks,
the text string associated with a keyword is restricted to the
Latin-1 character set plus the linefeed character. Text strings
in <a href="#11zTXt"><span class="chunk">zTXt</span></a> are
compressed into zlib datastreams using deflate compression (see
10.3: <a href='#10CompressionOtherUses'><span class="xref">Other
uses of compression</span></a>). The <a href="#11iTXt"><span
class="chunk">iTXt</span></a> chunk can be used to convey
characters outside the Latin-1 set. It uses the UTF-8 encoding of
UCS <a href="#2-ISO-10646-1"><span class="NormRef">[ISO/IEC
10646-1]</span></a> . There is an option to compress text strings
in the <a href="#11iTXt"><span class="chunk">iTXt</span></a>
chunk.</p>

<h4><a name="11tEXt">11.3.4.3 <span class="chunk">tEXt</span>
Textual data</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
116 69 88 116
</pre>

<p>Each <span class="chunk">tEXt</span> chunk contains a keyword
and a text string, in the format:</p>

<table class="Regular" summary=
"This table defines the tEXt chunk">
<tr>
<td class="Regular">Keyword</td>
<td class="Regular">1-79 bytes (character string)</td>
</tr>

<tr>
<td class="Regular">Null separator</td>
<td class="Regular">1 byte (null character)</td>
</tr>

<tr>
<td class="Regular">Text string</td>
<td class="Regular">0 or more bytes (character string)</td>
</tr>
</table>

<p>
The keyword and text string are separated by a zero byte (null
character). Neither the keyword nor the text string may contain a
null character.
The text string is <strong>not</strong> null-terminated (the length of
the chunk defines the ending). The text string may be of any
length from zero bytes up to the maximum permissible chunk size
less the length of the keyword and null character separator.</p>

<p>The keyword indicates the type of information represented by
the text string as described in 11.3.4.2: <a href=
"#11keywords"><span class="xref">Keywords and text
strings</span></a>.</p>

<p>Text is interpreted according to the Latin-1 character set <a
href="#2-ISO-8859-1"><span class=
"NormRef">[ISO-8859-1]</span></a>. The text string may contain
any Latin-1 character. Newlines in the text string should be
represented by a single linefeed character (decimal 10).
Characters other than those defined in Latin-1 plus the linefeed
character have no defined meaning in <span class="chunk">tEXt</span> chunks.
Text containing characters outside the repertoire of ISO/IEC
8859-1 should be encoded using the <a href="#11iTXt"><span class=
"chunk">iTXt</span></a> chunk.</p>

<h4><a name="11zTXt">11.3.4.4 <span class="chunk">zTXt</span>
Compressed textual data</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
122 84 88 116
</pre>

<p>The <span class="chunk">zTXt</span> and <a href=
"#11tEXt"><span class="chunk">tEXt</span></a> chunks are
semantically equivalent, but the <span class="chunk">zTXt</span>
chunk is recommended for storing large blocks of text.</p>

<p>A <span class="chunk">zTXt</span> chunk contains:</p>

<table class="Regular" summary=
"This table defines the zTXt chunk">
<tr>
<td class="Regular">Keyword</td>
<td class="Regular">1-79 bytes (character string)</td>
</tr>

<tr>
<td class="Regular">Null separator</td>
<td class="Regular">1 byte (null character)</td>
</tr>

<tr>
<td class="Regular">Compression method</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Compressed text datastream</td>
<td class="Regular">n bytes</td>
</tr>
</table>

<p>The keyword and null character are the same as in the <a href=
"#11tEXt"><span class="chunk">tEXt</span></a> chunk (see
11.3.4.3: <a href="#11tEXt"><span class="xref"><span class=
"chunk">tEXt</span> Textual data</span></a>). The keyword is not
compressed. The compression method entry defines the compression
method used. The only value defined in this International
Standard is 0 (deflate/inflate compression). Other values are
reserved for future standardization (see 4.9 <a href=
"#4Concepts.Registration"><span class="xref">Extension and
registration</span></a>). The compression method entry is
followed by the compressed text datastream that makes up the
remainder of the chunk. For compression method 0, this datastream
is a zlib datastream with deflate compression (see 10.3: <a href=
"#10CompressionOtherUses"><span class="xref">Other uses of
compression</span></a>). Decompression of this datastream yields
Latin-1 text that is identical to the text that would be stored
in an equivalent <a href="#11tEXt"><span class=
"chunk">tEXt</span></a> chunk.</p>

<h4><a name="11iTXt">11.3.4.5 <span class="chunk">iTXt</span>
International textual data</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
105 84 88 116
</pre>

<p>An <span class="chunk">iTXt</span> chunk contains:</p>

<table class="Regular" summary=
"This table defines the iTXt chunk">
<tr>
<td class="Regular">Keyword</td>
<td class="Regular">1-79 bytes (character string)</td>
</tr>

<tr>
<td class="Regular">Null separator</td>
<td class="Regular">1 byte (null character)</td>
</tr>

<tr>
<td class="Regular">Compression flag</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Compression method</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Language tag</td>
<td class="Regular">0 or more bytes (character string)</td>
</tr>

<tr>
<td class="Regular">Null separator</td>
<td class="Regular">1 byte (null character)</td>
</tr>

<tr>
<td class="Regular">Translated keyword</td>
<td class="Regular">0 or more bytes</td>
</tr>

<tr>
<td class="Regular">Null separator</td>
<td class="Regular">1 byte (null character)</td>
</tr>

<tr>
<td class="Regular">Text</td>
<td class="Regular">0 or more bytes</td>
</tr>
</table>

<p>The keyword is described in 11.3.4.2: <a href=
"#11keywords"><span class="xref">Keywords and text
strings</span></a>.</p>

<p>The compression flag is 0 for uncompressed text, 1 for
compressed text. Only the text field may be compressed. The
compression method entry defines the compression method used. The
only compression method defined in this International Standard is
0 (zlib datastream with deflate compression, see 10.3: <a href=
'#10CompressionOtherUses'><span class="xref">Other uses of
compression</span></a>). For uncompressed text, encoders shall
set the compression method to 0, and decoders shall ignore
it.</p>

<p>The language tag defined in <a href="#2-RFC-3066"><span class=
"NormRef">[RFC-3066]</span></a>
indicates the human language used by the translated keyword and
the text. Unlike the keyword, the language tag is
case-insensitive. It is an ISO 646.IRV:1991 <a href="#2-ISO-646"><span
class="NormRef">[ISO 646]</span></a> string consisting of
hyphen-separated words of 1-8 alphanumeric characters each (for example cn,
en-uk, no-bok, x-klingon, x-KlInGoN). If the first word is two or three
letters long, it is an ISO language code <a href=
"#2-ISO-639"><span class="NormRef">[ISO-639]</span></a>. If the
language tag is empty, the language is unspecified.</p>

<p>The translated keyword and text both use the UTF-8 encoding of
UCS <a href="#2-ISO-10646-1"><span class="NormRef">[ISO/IEC
10646-1]</span></a>, and neither shall contain a zero byte (null
character). The text, unlike other textual data in this chunk, is
not null-terminated; its length is derived from the chunk
length.</p>

<p>Line breaks should not appear in the translated keyword. In
the text, a newline should be represented by a single linefeed
character (decimal 10). The remaining control characters (1-9,
11-31, 127-159) are discouraged in both the translated keyword
and text. In UTF-8 there is a difference between the characters
128-159 (which are discouraged) and the bytes 128-159 (which are
often necessary).</p>

<p>The translated keyword, if not empty, should contain a
translation of the keyword into the language indicated by the
language tag, and applications displaying the keyword should
display the translated keyword in addition.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h3><a name="11addnlsiinfo">11.3.5 Miscellaneous
information</a></h3>

<h4><a name="11bKGD">11.3.5.1 <span class="chunk">bKGD</span>
Background colour</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
98 75 71 68
</pre>

<p>The <span class="chunk">bKGD</span> chunk specifies a default
background colour to present the image against. If there is any
other preferred background, either user-specified or part of a
larger page (as in a browser), the <span class=
"chunk">bKGD</span> chunk should be ignored. The <span class=
"chunk">bKGD</span> chunk contains:</p>

<table class="Regular" summary=
"This table defines the bKGD chunk">
<tr>
<th colspan="2">Colour types 0 and 4</th>
</tr>

<tr>
<td class="Regular">Greyscale</td>
<td class="Regular">2 bytes</td>
</tr>

<tr>
<th colspan="2">Colour types 2 and 6</th>
</tr>

<tr>
<td class="Regular">Red</td>
<td class="Regular">2 bytes</td>
</tr>

<tr>
<td class="Regular">Green</td>
<td class="Regular">2 bytes</td>
</tr>

<tr>
<td class="Regular">Blue</td>
<td class="Regular">2 bytes</td>
</tr>

<tr>
<th colspan="2">Colour type 3</th>
</tr>

<tr>
<td class="Regular">Palette index</td>
<td class="Regular">1 byte</td>
</tr>
</table>

<p>For colour type 3 (indexed-colour), the value is the palette
index of the colour to be used as background.</p>

<p>For colour types 0 and 4 (greyscale, greyscale with alpha),
the value is the grey level to be used as background in the range
0 to (2<sup>bitdepth</sup>)-1. For colour types 2 and 6
(truecolour, truecolour with alpha), the values are the colour to be
used as background, given as RGB
samples in the range 0 to (2<sup>bitdepth</sup>)-1. In each case,
for consistency, two bytes per sample are used regardless of the
image bit depth. If the image bit depth is less than 16, the
least significant bits are used and the others are 0.</p>

<h4><a name="11hIST">11.3.5.2 <span class="chunk">hIST</span>
Image histogram</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
104 73 83 84
</pre>

<p>The <span class="chunk">hIST</span> chunk contains a series of
two-byte (16-bit) unsigned integers:</p>

<table class="Regular" summary=
"This table defines the hIST chunk">
<tr>
<td class="Regular">Frequency</td>
<td class="Regular">2 bytes (unsigned integer)</td>
</tr>
<tr>
<td class="Regular">...etc...</td>
<td class="Regular">&nbsp;</td>
</tr>
</table>

<p>The <span class="chunk">hIST</span> chunk gives the
approximate usage frequency of each colour in the palette. A
histogram chunk can appear only when a <a href="#11PLTE"><span
class="chunk">PLTE</span></a> chunk appears. If a viewer is
unable to provide all the colours listed in the palette, the
histogram may help it decide how to choose a subset of the
colours for display.</p>

<p>There shall be exactly one
entry for each entry in the <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk. Each entry is proportional to the
fraction of pixels in the image that have that palette index; the
exact scale factor is chosen by the encoder.</p>

<p>Histogram entries are approximate, with the exception that a
zero entry specifies that the corresponding palette entry is not
used at all in the image. A histogram entry shall be nonzero if
there are any pixels of that colour.</p>

<p class="Note">NOTE When the palette is a suggested quantization
of a truecolour image, the histogram is necessarily approximate,
since a decoder may map pixels to palette entries differently
than the encoder did. In this situation, zero entries should not
normally appear, because any entry might be used.</p>

<h4><a name="11pHYs">11.3.5.3 <span class="chunk">pHYs</span>
Physical pixel dimensions</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
112 72 89 115
</pre>

<p>The <span class="chunk">pHYs</span> chunk specifies the
intended pixel size or aspect ratio for display of the image. It
contains:</p>

<table class="Regular" summary=
"This table defines the pHYs chunk">
<tr>
<td class="Regular">Pixels per unit, X axis</td>
<td class="Regular">4 bytes (PNG unsigned integer)</td>
</tr>

<tr>
<td class="Regular">Pixels per unit, Y axis</td>
<td class="Regular">4 bytes (PNG unsigned integer)</td>
</tr>

<tr>
<td class="Regular">Unit specifier</td>
<td class="Regular">1 byte</td>
</tr>
</table>

<p>The following values are defined for the unit specifier:</p>

<table class="Regular" summary=
"This table defines the allowed values for the unit specifier in the pHYs chunk">
<tr>
<td class="Regular">0</td>
<td class="Regular">unit is unknown</td>
</tr>

<tr>
<td class="Regular">1</td>
<td class="Regular">unit is the metre</td>
</tr>
</table>

<p>When the unit specifier is 0, the <span class=
"chunk">pHYs</span> chunk defines pixel aspect ratio only; the
actual size of the pixels remains unspecified.</p>

<p>If the <span class="chunk">pHYs</span> chunk is not present,
pixels are assumed to be square, and the physical size of each
pixel is unspecified.</p>

<h4><a name="11sPLT">11.3.5.4 <span class="chunk">sPLT</span>
Suggested palette</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
115 80 76 84
</pre>

<p>The <span class="chunk">sPLT</span> chunk contains:</p>

<table class="Regular" summary=
"This table defines the sPLT chunk">
<tr>
<td class="Regular">Palette name</td>
<td class="Regular">1-79 bytes (character string)</td>
</tr>

<tr>
<td class="Regular">Null separator</td>
<td class="Regular">1 byte (null character)</td>
</tr>

<tr>
<td class="Regular">Sample depth</td>
<td class="Regular">1 byte</td>
</tr>

<tr>
<td class="Regular">Red</td>
<td class="Regular">1 or 2 bytes</td>
</tr>

<tr>
<td class="Regular">Green</td>
<td class="Regular">1 or 2 bytes</td>
</tr>

<tr>
<td class="Regular">Blue</td>
<td class="Regular">1 or 2 bytes</td>
</tr>

<tr>
<td class="Regular">Alpha</td>
<td class="Regular">1 or 2 bytes</td>
</tr>

<tr>
<td class="Regular">Frequency</td>
<td class="Regular">2 bytes</td>
</tr>

<tr>
<td class="Regular">...etc...</td>
<td class="Regular">&nbsp;</td>
</tr>
</table>

<p>Each palette entry is six bytes or ten bytes containing five
unsigned integers (red, blue, green, alpha, and frequency).</p>

<p>There may be any number of entries. A PNG decoder determines
the number of entries from the length of the chunk remaining
after the sample depth byte. This shall be divisible by 6 if the
<span class="chunk">sPLT</span> sample depth is 8, or by 10 if
the <span class="chunk">sPLT</span> sample depth is 16. Entries
shall appear in decreasing order of frequency. There is no
requirement that the entries all be used by the image, nor that
they all be different.</p>

<p>The palette name can be any convenient name for referring to
the palette (for example "256 colour including Macintosh
default", "256 colour including Windows-3.1 default", "Optimal
512"). The palette name may aid the choice of the appropriate
suggested palette when more than one appears in a PNG
datastream.</p>

<p>The palette name is case-sensitive, and subject to the same
restrictions as the keyword parameter for the <a href=
"#11tEXt"><span class="chunk">tEXt</span></a> chunk. Palette
names shall contain only printable Latin-1 characters and spaces
(only character codes 32-126 and 161-255 decimal are allowed).
Leading, trailing, and consecutive spaces are not permitted.</p>

<p>The <span class="chunk">sPLT</span> sample depth shall be 8 or
16.</p>

<p>The red, green, blue, and alpha samples are either one or two
bytes each, depending on the <span class="chunk">sPLT</span>
sample depth, regardless of the image bit depth. The colour
samples are not premultiplied by alpha, nor are they
precomposited against any background. An alpha value of 0 means
fully transparent. An alpha value of 255 (when the <span class=
"chunk">sPLT</span> sample depth is 8) or 65535 (when the <span
class="chunk">sPLT</span> sample depth is 16) means fully opaque.
The <span class="chunk">sPLT</span> chunk may appear for any PNG
colour type. Entries in <span class="chunk">sPLT</span> use the
same gamma and chromaticity values as the PNG image, but may fall
outside the range of values used in the colour space of the PNG
image; for example, in a greyscale PNG image, each <span class=
"chunk">sPLT</span> entry would typically have equal red, green,
and blue values, but this is not required. Similarly, <span
class="chunk">sPLT</span> entries can have non-opaque alpha
values even when the PNG image does not use transparency.</p>

<p>Each frequency value is proportional to the fraction of 
the pixels in the image for which that palette entry
is the closest match in RGBA space, before the image has been composited against any
background. The exact scale factor is chosen by the PNG encoder;
it is recommended that the resulting range of individual values
reasonably fills the range 0 to 65535. A PNG encoder may
artificially inflate the frequencies for colours considered to be
"important", for example the colours used in a logo or the facial
features of a portrait. Zero is a valid frequency meaning that
the colour is "least important" or that it is rarely, if ever,
used. When all the frequencies are zero, they are meaningless,
that is to say, nothing may be inferred about the actual
frequencies with which the colours appear in the PNG image.</p>

<p>Multiple <span class="chunk">sPLT</span> chunks are permitted,
but each shall have a different palette name.</p>

<h3><a name="11timestampinfo">11.3.6 Time stamp
information</a></h3>

<h4><a name="11tIME">11.3.6.1 <span class="chunk">tIME</span>
Image last-modification time</a></h4>

<p>The four-byte chunk type field contains the decimal values</p>

<pre>
116 73 77 69
</pre>

<p>The <span class="chunk">tIME</span> chunk gives the time of
the last image modification (<strong>not</strong> the time of initial
image creation). It contains:</p>

<table class="Regular" summary=
"This table defines the tIME chunk">
<tr>
<td class="Regular">Year</td>
<td class="Regular">2 bytes (complete; for example, 1995, <strong>not</strong> 95)</td>
</tr>

<tr>
<td class="Regular">Month</td>
<td class="Regular">1 byte (1-12)</td>
</tr>

<tr>
<td class="Regular">Day</td>
<td class="Regular">1 byte (1-31)</td>
</tr>

<tr>
<td class="Regular">Hour</td>
<td class="Regular">1 byte (0-23)</td>
</tr>

<tr>
<td class="Regular">Minute</td>
<td class="Regular">1 byte (0-59)</td>
</tr>

<tr>
<td class="Regular">Second</td>
<td class="Regular">1 byte (0-60) (to allow for leap seconds)</td>
</tr>
</table>

<p>Universal Time (UTC) should be specified rather than local
time.</p>

<p>The <span class="chunk">tIME</span> chunk is intended for use
as an automatically-applied time stamp that is updated whenever
the image data are changed.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="12Encoders">12 PNG Encoders</a></h1>

<h2><a name="12Introduction">12.1 Introduction</a></h2>

<p>This clause gives requirements and recommendations for encoder
behaviour. A PNG encoder shall produce a PNG datastream from a
PNG image that conforms to the format specified in the preceding
clauses. Best results will usually be achieved by following the
additional recommendations given here.</p>

<h2><a name="12Encoder-gamma-handling">12.2 Encoder gamma
handling</a></h2>

<p>See Annex C: <a href="#C-GammaAppendix"><span class=
"xref">Gamma and chromaticity</span></a> for a brief introduction
to gamma issues.</p>

<p>PNG encoders capable of full colour management <a href=
"#G-ICC"><span class="bibref">[ICC]</span></a> will perform more
sophisticated calculations than those described here and may
choose to use the <a href="#11iCCP"><span class=
"chunk">iCCP</span></a> chunk. If it is known that the image
samples conform to the sRGB specification <a href=
"#2-IEC-61966-2-1"><span class="NormRef">[IEC
61966-2-1]</span></a>, encoders are strongly encouraged to write
the <a href="#11sRGB"><span class="chunk">sRGB</span></a> chunk
without performing additional gamma handling. In both cases it is
recommended that an appropriate <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> chunk be generated for use by PNG
decoders that do not recognize the <a href="#11iCCP"><span class=
"chunk">iCCP</span></a> chunk or <a href="#11sRGB"><span class=
"chunk">sRGB</span></a> chunk.</p>

<p>A PNG encoder has to determine:</p>

<!-- <ol start="1"> --><ol>
<li>what value to write in the <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> chunk;</li>

<li>how to transform the provided image samples  into the values
to be written in the PNG datastream.</li>
</ol>

<p>The value to write in the <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> chunk is that value which causes a PNG
decoder to behave in the desired way. See 13.13: <a class='Href'
href='#13Decoder-gamma-handling'>Decoder gamma handling</a>.</p>

<p>The transform to be applied depends on the nature of the image
samples and their precision. If the samples represent light
intensity in floating-point or high precision integer form
(perhaps from a computer graphics renderer), the encoder may
perform "gamma encoding" (applying a power function with exponent
less than 1) before quantizing the data to integer values for
inclusion in the PNG datastream. This results in fewer banding
artifacts at a given sample depth, or allows smaller samples
while retaining the same visual quality. An intensity level
expressed as a floating-point value in the range 0 to 1 can be
converted to a datastream image sample by:</p>

<p><tt>integer_sample =
floor((2<sup>sampledepth</sup>-1) * intensity<sup>encoding_exponent</sup>
+ 0.5)</tt></p>

<p>If the intensity in the equation is the desired output
intensity, the encoding exponent is the gamma value to be used in
the <a href="#11gAMA"><span class="chunk">gAMA</span></a>
chunk.</p>

<p>If the intensity available to the PNG encoder is the original
scene intensity, another transformation may be needed. There is
sometimes a requirement for the displayed image to have higher
contrast than the original source image. This corresponds to an
end-to-end transfer function from original scene to display
output with an exponent greater than 1. In this case:</p>

<pre>
gamma = encoding_exponent/end_to_end_exponent
</pre>

<p>If it is not known whether the conditions under which the
original image was captured or calculated warrant such a contrast
change, it may be assumed that the display intensities are
proportional to original scene intensities, i.e. the end-to-end
exponent is 1 and hence:</p>

<pre>
gamma = encoding_exponent
</pre>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p>If the image is being written to a datastream only, the
encoder is free to choose the encoding exponent. Choosing a value
that causes the gamma value in the <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> chunk to be 1/2.2 is often a reasonable
choice because it minimizes the work for a PNG decoder displaying
on a typical video monitor.</p>

<p>Some image renderers may simultaneously write the image to a
PNG datastream and display it on-screen. The displayed pixels
should be gamma corrected for the display system and viewing
conditions in use, so that the user sees a proper representation
of the intended scene.</p>

<p>If the renderer wants to write the displayed sample values to
the PNG datastream, avoiding a separate gamma encoding step for
the datastream, the renderer should approximate the transfer
function of the display system by a power function, and write the
reciprocal of the exponent into the <a href="#11gAMA"><span
class="chunk">gAMA</span></a> chunk. This will allow a PNG
decoder to reproduce what was displayed on screen for the
originator during rendering.</p>

<p>However, it is equally reasonable for a renderer to compute
displayed pixels appropriate for the display device, and to
perform separate gamma encoding for data storage and
transmission, arranging to have a value in the <a href=
"#11gAMA"><span class="chunk">gAMA</span></a> chunk more
appropriate to the future use of the image.</p>

<p>Computer graphics renderers often do not perform gamma
encoding, instead making sample values directly proportional to
scene light intensity. If the PNG encoder receives sample values
that have already been quantized into integer values, there is no
point in doing gamma encoding on them; that would just result in
further loss of information. The encoder should just write the
sample values to the PNG datastream. This does not imply that the
<a href="#11gAMA"><span class="chunk">gAMA</span></a> chunk
should contain a gamma value of 1.0 because the desired
end-to-end transfer function from scene intensity to display
output intensity is not necessarily linear. However, the desired
gamma value is probably not far from 1.0. It may depend on
whether the scene being rendered is a daylight scene or an indoor
scene, etc.</p>

<p>When the sample values come directly from a piece of hardware,
the correct <a href="#11gAMA"><span class="chunk">gAMA</span></a>
value can, in principle, be inferred from the transfer function
of the hardware and lighting conditions of the scene. In the case
of video digitizers ("frame grabbers"), the samples are probably
in the sRGB colour space, because the sRGB specification was
designed to be compatible with modern video standards. Image
scanners are less predictable. Their output samples may be
proportional to the input light intensity since CCD sensors
themselves are linear, or the scanner hardware may have already
applied a power function designed to compensate for dot gain in
subsequent printing (an exponent of about 0.57), or the scanner
may have corrected the samples for display on a monitor. It may
be necessary to refer to the scanner's manual or to scan a
calibrated target in order to determine the characteristics of a
particular scanner. It should be remembered that gamma relates
samples to desired display output, not to scanner input.</p>

<p>Datastream format converters generally should not attempt to
convert supplied images to a different gamma. The data should be
stored in the PNG datastream without conversion, and the gamma
value should be deduced from information in the source datastream
if possible. Gamma alteration at datastream conversion time
causes re-quantization of the set of intensity levels that are
represented, introducing further roundoff error with little
benefit. It is almost always better to just copy the sample
values intact from the input to the output file.</p>

<p>If the source datastream describes the gamma characteristics
of the image, a datastream converter is strongly encouraged to
write a <a href="#11gAMA"><span class="chunk">gAMA</span></a>
chunk. Some datastream formats specify the display exponent (the
exponent of the function which maps image samples to display
output rather than the other direction). If the source file's
gamma value is greater than 1.0, it is probably a display
exponent, and the reciprocal of this value should be used for the
PNG gamma value. If the source file format records the
relationship between image samples and a quantity other than
display output, it will be more complex than this to deduce the
PNG gamma value.</p>

<p>If a PNG encoder or datastream converter knows that the image
has been displayed satisfactorily using a display system whose
transfer function can be approximated by a power function with
exponent <tt>display_exponent</tt>, the image can be marked as
having the gamma value:</p>

<pre>
gamma = 1/display_exponent
</pre>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p>It is better to write a <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> chunk with a value that is approximately
correct than to omit the chunk and force PNG decoders to guess an
approximate gamma. If a PNG encoder is unable to infer the gamma
value, it is preferable to omit the <a href="#11gAMA"><span
class="chunk">gAMA</span></a> chunk. If a guess has to be made
this should be left to the PNG decoder.</p>

<p>Gamma does not apply to alpha samples; alpha is always
represented linearly.</p>

<p>See also 13.13: <a href="#13Decoder-gamma-handling"><span
class="xref">Decoder gamma handling</span></a>.</p>

<h2><a name="12Encoder-colour-handling">12.3 Encoder colour
handling</a></h2>

<p>See Annex C: <a href="#C-GammaAppendix"><span class=
"xref">Gamma and chromaticity</span></a> for references to colour
issues.</p>

<p>PNG encoders capable of full colour management <a href=
"#G-ICC"><span class="bibref">[ICC]</span></a> will perform more
sophisticated calculations than those described here and may
choose to use the <a href="#11iCCP"><span class=
"chunk">iCCP</span></a> chunk. If it is known that the image
samples conform to the sRGB specification <a href=
"#2-IEC-61966-2-1"><span class="NormRef">[IEC
61966-2-1]</span></a>, PNG encoders are strongly encouraged to
use the <a href="#11sRGB"><span class="chunk">sRGB</span></a>
chunk.</p>

<p>If it is possible for the encoder to determine the
chromaticities of the source display primaries, or to make a
strong guess based on the origin of the image, or the hardware
running it, the encoder is strongly encouraged to output the <a
href="#11cHRM"><span class="chunk">cHRM</span></a> chunk. If this
is done, the <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> chunk should also be written; decoders
can do little with a <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunk if the <a href="#11gAMA"><span
class="chunk">gAMA</span></a> chunk is missing.</p>

<p>There are a number of recommendations and standards for
primaries and white points, some of which are linked to
particular technologies, for example the CCIR 709 standard <a
href="#G-ITU-R-BT709"><span class=
"bibref">[ITU-R-BT709]</span></a> and the SMPTE-C standard <a
href="#G-SMPTE-170M"><span class=
"bibref">[SMPTE-170M]</span></a>.</p>

<p>There are three cases that need to be considered:</p>

<ol>
<li>the encoder is part of the generation system;</li>

<li>the source image is captured by a camera or scanner;</li>

<li>the PNG datastream was generated by translation from some
other format.</li>
</ol>

<!--  deleted - comment PDG 31<p>Scanners that produce PNG datastreams as output should insert
the filter chromaticities into a <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunk.</p>-->

<p>In the case of hand-drawn or digitally edited images, it is
necessary to determine what monitor they were viewed on when
being produced. Many image editing programs allow the type of
monitor being used to be specified. This is often because they
are working in some device-independent space internally. Such
programs have enough information to write valid <a href=
"#11cHRM"><span class="chunk">cHRM</span></a> and <a href=
"#11gAMA"><span class="chunk">gAMA</span></a> chunks, and are
strongly encouraged to do so automatically.</p>

<p>If the encoder is compiled as a portion of a computer image
renderer that performs full-spectral rendering, the monitor
values that were used to convert from the internal
device-independent colour space to RGB should be written into the
<a href="#11cHRM"><span class="chunk">cHRM</span></a> chunk. Any
colours that are outside the gamut of the chosen RGB device
should be mapped to be within the gamut; PNG does not store
out-of-gamut colours.</p>

<p>If the computer image renderer performs calculations directly
in device-dependent RGB space, a <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunk should not be written unless the
scene description and rendering parameters have been adjusted for
a particular monitor. In that case, the data for that monitor
should be used to construct a <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunk.</p>

<p>A few image formats store calibration information, which can
be used to fill in the <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunk. For example, TIFF 6.0 files <a
href="#G-TIFF-6.0"><span class="bibref">[TIFF-6.0]</span></a> can
optionally store calibration information, which if present should
be used to construct the <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunk.</p>

<p>Video created with recent video equipment probably uses the
CCIR 709 primaries and D65 white point <a href=
"#G-ITU-R-BT709"><span class="bibref">[ITU-R-BT709]</span></a>,
which are given in <a href="#12-table121"><span class=
"tabref">Table 12.1</span></a>.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<table class="Regular" summary=
"CCIR 709 primaries and D65 whitepoint">
<caption><a name="12-table121"><b>Table 12.1 &mdash; CCIR 709
primaries and D65 whitepoint</b></a></caption>

<tr>
<th>&nbsp;</th>
<th>R</th>
<th>G</th>
<th>B</th>
<th>White</th>
</tr>

<tr>
<td class="Regular">x</td>
<td class="Regular">0.640</td>
<td class="Regular">0.300</td>
<td class="Regular">0.150</td>
<td class="Regular">0.3127</td>
</tr>

<tr>
<td class="Regular">y</td>
<td class="Regular">0.330</td>
<td class="Regular">0.600</td>
<td class="Regular">0.060</td>
<td class="Regular">0.3290</td>
</tr>
</table>

<p>An older but still very popular video standard is SMPTE-C <a
href="#G-SMPTE-170M"><span class="bibref">[SMPTE-170M]</span></a>
given in <a href="#12-table122"><span class="tabref">Table
12.2</span></a>.</p>

<table class="Regular" summary=
"CSMPTE-C video standard">
<caption><a name="12-table122"><b>Table 12.2 &mdash; SMPTE-C
video standard</b></a></caption>

<tr>
<th>&nbsp;</th>
<th>R</th>
<th>G</th>
<th>B</th>
<th>White</th>
</tr>

<tr>
<td class="Regular">x</td>
<td class="Regular">0.630</td>
<td class="Regular">0.310</td>
<td class="Regular">0.155</td>
<td class="Regular">0.3127</td>
</tr>

<tr>
<td class="Regular">y</td>
<td class="Regular">0.340</td>
<td class="Regular">0.595</td>
<td class="Regular">0.070</td>
<td class="Regular">0.3290</td>
</tr>
</table>

<p>It is <strong>not</strong> recommended that datastream format
converters attempt to convert supplied images to a different RGB
colour space. The data should be stored in the PNG datastream
without conversion, and the source primary chromaticities should
be recorded if they are known. Colour space transformation at
datastream conversion time is a bad idea because of gamut
mismatches and rounding errors. As with gamma conversions, it is
better to store the data losslessly and incur at most one
conversion when the image is finally displayed.</p>

<p>See also 13.14: <a href="#13Decoder-colour-handling"><span
class="xref">Decoder colour handling</span></a>.</p>

<h2><a name="12Alpha-channel-creation">12.4 Alpha channel
creation</a></h2>

<p>The alpha channel can be regarded either as a mask that
temporarily hides transparent parts of the image, or as a means
for constructing a non-rectangular image. In the first case, the
colour values of fully transparent pixels should be preserved for
future use. In the second case, the transparent pixels carry no
useful data and are simply there to fill out the rectangular
image area required by PNG. In this case, fully transparent
pixels should all be assigned the same colour value for best
compression.</p>

<p>Image authors should keep in mind the possibility that a
decoder will not support transparency control in full (see 13.16:
<a href="#13Alpha-channel-processing"><span class="xref">Alpha
channel processing</span></a>). Hence, the colours assigned to
transparent pixels should be reasonable background colours
whenever feasible.</p>

<p>For applications that do not require a full alpha channel, or
cannot afford the price in compression efficiency, the <a href=
"#11tRNS"><span class="chunk">tRNS</span></a> transparency chunk
is also available.</p>

<p>If the image has a known background colour, this colour should
be written in the <a href="#11bKGD"><span class=
"chunk">bKGD</span></a> chunk. Even decoders that ignore
transparency may use the <a href="#11bKGD"><span class=
"chunk">bKGD</span></a> colour to fill unused screen area.</p>

<p>If the original image has premultiplied (also called
"associated") alpha data, it can be converted to PNG's
non-premultiplied format by dividing each sample value by the
corresponding alpha value, then multiplying by the maximum value
for the image bit depth, and rounding to the nearest integer. In
valid premultiplied data, the sample values never exceed their
corresponding alpha values, so the result of the division should
always be in the range 0 to 1. If the alpha value is zero, output
black (zeroes).</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="12Sample-depth-scaling">12.5 Sample depth
scaling</a></h2>

<p>When encoding input samples that have a sample depth that
cannot be directly represented in PNG, the encoder shall scale
the samples up to a sample depth that is allowed by PNG. The most
accurate scaling method is the linear equation:</p>

<pre>
output = floor((input * MAXOUTSAMPLE / MAXINSAMPLE) + 0.5)
</pre>

<p>where the input samples range from 0 to <tt>MAXINSAMPLE</tt>
and the outputs range from 0 to <tt>MAXOUTSAMPLE</tt> (which is
2<sup>sampledepth</sup>-1).</p>

<p>A close approximation to the linear scaling method is achieved
by "left bit replication", which is shifting the valid bits to
begin in the most significant bit and repeating the most
significant bits into the open bits. This method is often faster
to compute than linear scaling.</p>

<p>EXAMPLE Assume that 5-bit samples are being scaled up to 8
bits. If the source sample value is 27 (in the range from 0-31),
then the original bits are:</p>

<pre>
   4 3 2 1 0
   ---------
   1 1 0 1 1
</pre>

<p>Left bit replication gives a value of 222:</p>

<pre>
   7 6 5 4 3  2 1 0
   ----------------
   1 1 0 1 1  1 1 0
   |=======|  |===|
       |      Leftmost Bits Repeated to Fill Open Bits
       |
   Original Bits
</pre>

<p>which matches the value computed by the linear equation. Left
bit replication usually gives the same value as linear scaling,
and is never off by more than one.</p>

<p>A distinctly less accurate approximation is obtained by simply
left-shifting the input value and filling the low order bits with
zeroes. This scheme cannot reproduce white exactly, since it does
not generate an all-ones maximum value; the net effect is to
darken the image slightly. This method is not recommended in
general, but it does have the effect of improving compression,
particularly when dealing with greater-than-8-bit sample depths.
Since the relative error introduced by zero-fill scaling is small
at high sample depths, some encoders may choose to use it.
Zero-fill shall <strong>not</strong> be used for alpha channel
data, however, since many decoders will treat alpha values of all
zeroes and all ones as special cases. It is important to
represent both those values exactly in the scaled data.</p>

<p>When the encoder writes an <a href="#11sBIT"><span class=
"chunk">sBIT</span></a> chunk, it is required to do the scaling
in such a way that the high-order bits of the stored samples
match the original data. That is, if the <a href="#11sBIT"><span
class="chunk">sBIT</span></a> chunk specifies a sample depth of
S, the high-order S bits of the stored data shall agree with the
original S-bit data values. This allows decoders to recover the
original data by shifting right. The added low-order bits are not
constrained. All the above scaling methods meet this
restriction.</p>

<p>When scaling up source image data, it is recommended that the
low-order bits be filled consistently for all samples; that is,
the same source value should generate the same sample value at
any pixel position. This improves compression by reducing the
number of distinct sample values. This is not a mandatory
requirement, and some encoders may choose not to follow it. For
example, an encoder might instead dither the low-order bits,
improving displayed image quality at the price of increasing file
size.</p>

<p>In some applications the original source data may have a range
that is not a power of 2. The linear scaling equation still works
for this case, although the shifting methods do not. It is
recommended that an <a href="#11sBIT"><span class=
"chunk">sBIT</span></a> chunk not be written for such images,
since <a href="#11sBIT"><span class="chunk">sBIT</span></a>
suggests that the original data range was exactly
0..2<sup>S</sup>-1.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="12Suggested-palettes">12.6 Suggested
palettes</a></h2>

<p>Suggested palettes may appear as <a href="#11sPLT"><span
class="chunk">sPLT</span></a> chunks in any PNG datastream, or as
a <a href="#11PLTE"><span class="chunk">PLTE</span></a> chunk in
truecolour PNG datastreams. In either case, the suggested palette
is not an essential part of the image data, but it may be used to
present the image on indexed-colour display hardware. Suggested
palettes are of no interest to viewers running on truecolour
hardware.</p>

<p>When an <a href="#11sPLT"><span class="chunk">sPLT</span></a>
chunk is used to provide a suggested palette, it is recommended
that the encoder use the frequency fields to indicate the
relative importance of the palette entries, rather than leave
them all zero (meaning undefined). The frequency values are most
easily computed as "nearest neighbour" counts, that is, the
approximate usage of each RGBA palette entry if no dithering is
applied. (These counts will often be available "for free" as a
consequence of developing the suggested palette.) Because the
suggested palette includes transparency information, it should be
computed for the uncomposited image.</p>

<p>Even for indexed-colour images, <a href="#11sPLT"><span class=
"chunk">sPLT</span></a> can be used to define alternative reduced
palettes for viewers that are unable to display all the colours
present in the <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk.
If the <a href="#11PLTE"><span class="chunk">PLTE</span></a>
chunk appears without the <a href="#11bKGD"><span class=
"chunk">bKGD</span></a> chunk in an image of colour type 6, the
circumstances under which the palette was computed are
unspecified.</p>


<p>An older method for including a suggested palette in a
truecolour PNG datastream uses the <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk. If this method is used, the
histogram (frequencies) should appear in a separate <a href=
"#11hIST"><span class="chunk">hIST</span></a> chunk. The <a href=
"#11PLTE"><span class="chunk">PLTE</span></a> chunk does not
include transparency information. Hence for images of colour type
6 (truecolour with alpha), it is recommended that a <a href=
"#11bKGD"><span class="chunk">bKGD</span></a> chunk appear and
that the palette and histogram be computed with reference to the
image as it would appear after compositing against the specified
background colour. This definition is necessary to ensure that
useful palette entries are generated for pixels having fractional
alpha values. The resulting palette will probably be useful only
to viewers that present the image against the same background
colour. It is recommended that PNG editors delete or recompute
the palette if they alter or remove the <a href="#11bKGD"><span
class="chunk">bKGD</span></a> chunk in an image of colour type
6.</p>

<p>For images of colour type 2 (truecolour), it is recommended
that the <a href="#11PLTE"><span class="chunk">PLTE</span></a>
and <a href="#11hIST"><span class="chunk">hIST</span></a> chunks
be computed with reference to the RGB data only, ignoring any
transparent-colour specification. If the datastream uses
transparency (has a <a href="#11tRNS"><span class=
"chunk">tRNS</span></a> chunk), viewers can easily adapt the
resulting palette for use with their intended background colour
(see 13.17: <a href=
"#13Histogram-and-suggested-palette-usage"><span class="xref">
Histogram and suggested palette usage</span></a>).
</p>

<p>For providing suggested palettes, 
the <a href="#11sPLT"><span class="chunk">sPLT</span></a>
chunk is more flexible than the <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk in
the following ways:</p>

<!-- <ol start="1"> --><ol>
<li>With <a href="#11sPLT"><span class="chunk">sPLT</span></a>
multiple suggested palettes may be provided. A PNG decoder may
choose an appropriate palette based on name or number of
entries.</li>

<li>In a PNG datastream of colour type 6 (truecolour with alpha
channel), the <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk represents a palette already
composited against the <a href="#11bKGD"><span class=
"chunk">bKGD</span></a> colour, so it is useful only for display
against that background colour. The <a href="#11sPLT"><span
class="chunk">sPLT</span></a> chunk provides an uncomposited
palette, which is useful for display against backgrounds chosen
by the PNG decoder.</li>

<li>Since the <a href="#11sPLT"><span class=
"chunk">sPLT</span></a> chunk is an ancillary chunk, a PNG editor
may add or modify suggested palettes without being forced to
discard unknown unsafe-to-copy chunks.</li>

<li>Whereas the <a href="#11sPLT"><span class=
"chunk">sPLT</span></a> chunk is allowed in PNG datastreams for
colour types 0, 3, and 4 (greyscale and indexed), the <a href=
"#11PLTE"><span class="chunk">PLTE</span></a> chunk cannot be
used to provide reduced palettes in these cases.</li>

<li>More than 256 entries may appear in the <a href=
"#11sPLT"><span class="chunk">sPLT</span></a> chunk.</li>
</ol>

<p>A PNG encoder that uses the <a href="#11sPLT"><span class=
"chunk">sPLT</span></a> chunk may choose to write a suggested
palette represented by <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> and <a href="#11hIST"><span class=
"chunk">hIST</span></a> chunks as well, for compatibility with
decoders that do not recognize the <a href="#11sPLT"><span class=
"chunk">sPLT</span></a> chunk.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="12Interlacing">12.7 Interlacing</a></h2>

<p>This International Standard defines two interlace methods,
one of which is no interlacing. Interlacing provides a convenient
basis from which decoders can progressively display an image, as
described in 13.8: <a href="#13Progressive-display"><span class=
"xref">Interlacing and progressive display</span></a>.</p>

<h2><a name="12Filter-selection">12.8 Filter selection</a></h2>

<p>For images of colour type 3 (indexed-colour), filter type 0
(None) is usually the most effective. Colour images with 256 or
fewer colours should almost always be stored in indexed-colour
format; truecolour format is likely to be much larger.</p>

<p>Filter type 0 is also recommended for images of bit depths
less than 8. For low-bit-depth greyscale images, in rare cases,
better compression may be obtained by first expanding the image
to 8-bit representation and then applying filtering.</p>

<p>For truecolour and greyscale images, any of the five filters
may prove the most effective. If an encoder uses a fixed filter,
the Paeth filter is most likely to be the best.</p>

<p>For best compression of truecolour and greyscale images,
the recommended approach is
adaptive filtering in which a filter is
chosen for each scanline. The following simple heuristic has
performed well in early tests: compute the output scanline using
all five filters, and select the filter that gives the smallest
sum of absolute values of outputs. (Consider the output bytes as
signed differences for this test.) This method usually
outperforms any single fixed filter choice. However, it is likely
that better heuristics will be found as more experience is
gained with PNG.</p>

<p>Filtering according to these recommendations is effective in
conjunction with either of the two interlace methods defined in
this International Standard.</p>

<h2><a name="12Compression">12.9 Compression</a></h2>

<p>The encoder may divide the compressed datastream into <a href=
"#11IDAT"><span class="chunk">IDAT</span></a> chunks however it
wishes. (Multiple <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunks are allowed so that encoders may
work in a fixed amount of memory; typically the chunk size will
correspond to the encoder's buffer size.) A PNG datastream in
which each <a href="#11IDAT"><span class="chunk">IDAT</span></a>
chunk contains only one data byte is valid, though remarkably
wasteful of space. (Zero-length <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunks are also valid, though even more
wasteful.)</p>

<h2><a name="12Text-chunk-processing">12.10 Text chunk
processing</a></h2>

<p>A nonempty keyword shall be provided for each text chunk. The
generic keyword "Comment" can be used if no better description of
the text is available. If a user-supplied keyword is used,
encoders should check that it meets the restrictions on
keywords.</p>

<p>For the <a href="#11tEXt"><span class="chunk">tEXt</span></a>
and <a href="#11zTXt"><span class="chunk">zTXt</span></a> chunks,
PNG text strings are expected to use the Latin-1 character set.
Encoders should avoid storing characters that are not defined in
Latin-1, and should provide character code remapping if the local
system's character set is not Latin-1. The <a href=
"#11iTXt"><span class="chunk">iTXt</span></a> chunk provides
support for international text, represented using the UTF-8
encoding of UCS. Encoders should discourage the creation of
single lines of text longer than 79 characters, in order to
facilitate easy reading. It is recommended that text items less
than 1024 bytes in size should be output using uncompressed 
text chunks. It is
recommended that the basic title and author keywords be output
using uncompressed text chunks. 
Placing large text chunks after the
image data (after the <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunks) can speed up image display in
some situations, as the decoder will decode the image data first.
It is recommended that small text chunks, such as the image
title, appear before the <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunks.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h2><a name="12Chunk-processing">12.11 Chunking</a></h2>

<h3><a name="12Use-of-private-chunks">12.11.1 Use of private
chunks</a></h3>

<p>
Chunk types are classified as public or private depending on bit 5
of the second byte (the private bit), and classified as critical or
ancillary depending on bit 5 of the first byte (the ancillary bit).
See 5.4: <a href=
"#5Chunk-naming-conventions"><span class="xref">Chunk naming
conventions</span></a>.
</p>

<p>Applications can use PNG private chunks to carry information
that need not be understood by other applications. Such chunks
shall be given private chunk types,
to ensure that they can never conflict
with any future public chunk definition. However, there is no
guarantee that some other application will not use the same
private chunk type. If a private chunk type is used, it is
prudent to store additional identifying information at the
beginning of the chunk data.</p>

<p>An ancillary chunk type, not a critical chunk type, should be
used for all private chunks that store information that is not
absolutely essential to view the image. Creation of private
critical chunks is discouraged because PNG datastreams containing
such chunks are not portable. Such chunks should not be used in
publicly available software or datastreams. If private critical
chunks are essential for an application, it is recommended that
one appear near the start of the datastream, so that a standard
decoder need not read very far before discovering that it cannot
handle the datastream.</p>

<p>If other organizations need to understand a new chunk type, it
should be submitted to the Registration Authority (see 4.9: <a
href="#4Concepts.Registration"><span class="xref">Extension and
registration</span></a>). A proposed public chunk type
shall not be used in publicly available software or
datastreams until registration has been approved.</p>

<p>If an ancillary chunk contains textual information that might
be of interest to a human user, a special chunk type should not
be defined for it. Instead a <a href="#11tEXt"><span class=
"chunk">tEXt</span></a> chunk should be used and a suitable
keyword defined. The information will then be available to other
users.</p>

<p>Keywords in <a href="#11tEXt"><span class=
"chunk">tEXt</span></a> chunks should be reasonably
self-explanatory, since the aim is to let other users understand
what the chunk contains. If generally useful, new keywords should
be registered with the Registration Authority (see 4.9: <a href=
"#4Concepts.Registration"><span class="xref">Extension and
registration</span></a>). However, it is permissible to use
keywords without registering them first.</p>

<h3><a name="12Private-type-and-method-codes">12.11.2 Private
type and method codes</a></h3>

<p>This specification defines the meaning of only some of the
possible values of some fields. For example, only compression
method 0 and filter types 0 through 4 are defined in this
International Standard. Numbers greater than 127 shall be used
when inventing experimental or private definitions of values for
any of these fields. Numbers below 128 are reserved for possible
public extensions of this specification through future
standardization (see 4.9 <a href="#4Concepts.Registration"><span
class="xref">Extension and registration</span></a>). The use of
private type codes may render a datastream unreadable by standard
decoders. Such codes are strongly discouraged except for
experimental purposes, and should not appear in publicly
available software or datastreams.</p>

<h3><a name="12Ancillary">12.11.3 Ancillary chunks</a></h3>

<p>All ancillary chunks are optional, encoders need not write
them. However, encoders are encouraged to write the standard
ancillary chunks when the information is available.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="13Decoders">13 PNG decoders and viewers</a></h1>

<h2><a name="13Introduction">13.1 Introduction</a></h2>

<p>This clause gives some requirements and recommendations for PNG
decoder behaviour and viewer behaviour. A viewer presents the
decoded PNG image to the user. Since viewer and decoder behaviour
are closely connected, decoders and viewers are treated together
here. The only absolute requirement on a PNG decoder is that it
successfully reads any datastream conforming to the format
specified in the preceding chapters. However, best results will
usually be achieved by following these additional
recommendations.</p>

<p>PNG decoders shall support all valid combinations of bit
depth, colour type, compression method, filter method, and
interlace method that are explicitly defined in this
International Standard.</p>

<p>All ancillary chunks are optional; decoders may ignore them.
However, decoders are encouraged to interpret these chunks when
appropriate and feasible.</p>

<h2><a name="13Decoders.Errors">13.2 Error handling</a></h2>

<p>Errors in a PNG datastream will fall into two general classes,
transmission errors and syntax errors (see <a href=
"#4Concepts.Errors"><span class="xref">4.8 Error
handling</span></a>).</p>

<p>Examples of transmission errors are transmission in "text" or
"ascii" mode, in which byte codes 13 and/or 10 may be added,
removed, or converted throughout the datastream; unexpected
termination, in which the datastream is truncated; or a physical
error on a storage device, in which one or more blocks (typically
512 bytes each) will have garbled or random values. Some examples
of syntax errors are an invalid value for a row filter, an
invalid compression method, an invalid chunk length, the absence
of a <a href="#11PLTE"><span class="chunk">PLTE</span></a> chunk
before the first <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunk in an indexed image, or the
presence of multiple <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> chunks. A PNG decoder should handle
errors as follows:</p>

<!-- <ol start="1"> --><ol>
<li>Detect errors as early as possible using the PNG signature
bytes and CRCs on each chunk. Decoders should verify that all
eight bytes of the PNG signature are correct. A decoder can
have additional confidence in the datastream's integrity if the
next eight bytes begin an <a href="#11IHDR"><span class=
"chunk">IHDR</span></a> chunk with the correct chunk length. A
CRC should be checked before processing the chunk data. Sometimes
this is impractical, for example when a streaming PNG decoder is
processing a large <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunk. In this case the CRC should be
checked when the end of the chunk is reached.</li>

<li>Recover from an error, if possible; otherwise fail
gracefully. Errors that have little or no effect on the
processing of the image may be ignored, while those that affect
critical data shall be dealt with in a manner appropriate to the
application.</li>

<li>Provide helpful messages describing errors, including
recoverable errors.</li>
</ol>

<p>Three classes of PNG chunks are relevant to this philosophy.
For the purposes of this classification, an "unknown chunk" is
either one whose type was genuinely unknown to the decoder's
author, or one that the author chose to treat as unknown, because
default handling of that chunk type would be sufficient for the
program's purposes. Other chunks are called "known chunks". Given
this definition, the three classes are as follows:</p>

<!-- <ol start="4"> --><ol>
<li>known chunks, which necessarily includes all of the critical
chunks defined in this International Standard (<a href=
"#11IHDR"><span class="chunk">IHDR</span></a>, <a href=
"#11PLTE"><span class="chunk">PLTE</span></a>, <a href=
"#11IDAT"><span class="chunk">IDAT</span></a>, <a href=
"#11IEND"><span class="chunk">IEND</span></a>)</li>

<li>unknown critical chunks (bit 5 of the first byte of the chunk
type is 0)</li>

<li>unknown ancillary chunks (bit 5 of the first byte of the
chunk type is 1)</li>
</ol>

<p>See 5.4: <a href="#5Chunk-naming-conventions"><span class=
"xref">Chunk naming conventions</span></a> for a full description
of chunk naming conventions.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p>PNG chunk types are marked "critical" or "ancillary" according
to whether the chunks are critical for the purpose of extracting
a viewable image (as with <a href="#11IHDR"><span class=
"chunk">IHDR</span></a>, <a href="#11PLTE"><span class=
"chunk">PLTE</span></a>, and <a href="#11IDAT"><span class=
"chunk">IDAT</span></a>) or critical to understanding the
datastream structure (as with <a href="#11IEND"><span class=
"chunk">IEND</span></a>). This is a specific kind of criticality
and one that is not necessarily relevant to every conceivable
decoder. For example, a program whose sole purpose is to extract
text annotations (for example, copyright information) does not
require a viewable image. Another decoder might consider the <a
href="#11tRNS"><span class="chunk">tRNS</span></a> and <a href=
"#11gAMA"><span class="chunk">gAMA</span></a> chunks essential to
its proper execution.</p>

<p>Syntax errors always involve known chunks because syntax
errors in unknown chunks cannot be detected. The PNG decoder has
to determine whether a syntax error is fatal (unrecoverable) or
not, depending on its requirements and the situation. For
example, most decoders can ignore an invalid <a href=
"#11IEND"><span class="chunk">IEND</span></a> chunk; a
text-extraction program can ignore the absence of <a href=
"#11IDAT"><span class="chunk">IDAT</span></a>; an image viewer
cannot recover from an empty <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk in an indexed image but it can
ignore an invalid <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk in a truecolour image; and a
program that extracts the alpha channel can ignore an invalid <a
href="#11gAMA"><span class="chunk">gAMA</span></a> chunk, but may
consider the presence of two <a href="#11tRNS"><span class=
"chunk">tRNS</span></a> chunks to be a fatal error. Anomalous
situations other than syntax errors shall be treated as
follows:</p>

<!-- <ol start="7"> --><ol>
<li>Encountering an unknown ancillary chunk is never an error.
The chunk can simply be ignored.</li>

<li>Encountering an unknown critical chunk is a fatal condition
for any decoder trying to extract the image from the datastream.
A decoder that ignored a critical chunk could not know whether
the image it extracted was the one intended by the encoder.</li>

<li>A PNG signature mismatch, a CRC mismatch, or an unexpected
end-of-stream indicates a corrupted datastream, and may be
regarded as a fatal error. A decoder could try to salvage
something from the datastream, but the extent of the damage will
not be known.</li>
</ol>

<p>When a fatal condition occurs, the decoder should fail
immediately, signal an error to the user if appropriate, and
optionally continue displaying any image data already visible to
the user (i.e. "fail gracefully"). The application as a whole
need not terminate.</p>

<p>When a non-fatal error occurs, the decoder should signal a
warning to the user if appropriate, recover from the error, and
continue processing normally.</p>

<p>Decoders that do not compute CRCs should interpret apparent
syntax errors as indications of corruption (see also 13.3: <a
href="#13Error-checking"><span class="xref">Error
checking</span></a>).</p>

<p>Errors in compressed chunks (<a href="#11IDAT"><span class=
"chunk">IDAT</span></a>, <a href="#11zTXt"><span class=
"chunk">zTXt</span></a>, <a href="#11iTXt"><span class=
"chunk">iTXt</span></a>, <a href="#11iCCP"><span class=
"chunk">iCCP</span></a>) could lead to buffer overruns.
Implementors of deflate decompressors should guard against this
possibility.</p>

<h2><a name="13Error-checking">13.3 Error checking</a></h2>

<p>The PNG error handling philosophy is described in 13.2: <a
href="#13Decoders.Errors"><span class="xref">Error
handling</span></a>.</p>

<p>Unknown chunk types shall be handled as described in 5.4: <a
href="#5Chunk-naming-conventions"><span class="xref">Chunk naming
conventions</span></a>. An unknown chunk type is <strong>not</strong> to
be treated as an error unless it is a critical chunk.</p>

<p>The chunk type can be checked for plausibility by seeing
whether all four bytes are in the range codes 65-90 and 97-122
(decimal); note that this need be done only for unrecognized
chunk types. If the total datastream size is known (from file
system information, HTTP protocol, etc), the chunk length can be
checked for plausibility as well. If CRCs are not checked,
dropped/added data bytes or an erroneous chunk length can cause
the decoder to get out of step and misinterpret subsequent data
as a chunk header.</p>

<p>For known-length chunks, such as <a href="#11IHDR"><span
class="chunk">IHDR</span></a>, decoders should treat an
unexpected chunk length as an error. Future extensions to this
specification will not add new fields to existing chunks;
instead, new chunk types will be added to carry new
information.</p>

<p>Unexpected values in fields of known chunks (for example, an
unexpected compression method in the <a href="#11IHDR"><span
class="chunk">IHDR</span></a> chunk) shall be checked for and
treated as errors. However, it is recommended that unexpected
field values be treated as fatal errors only in <strong>critical</strong>
chunks. An unexpected value in an ancillary chunk can be handled
by ignoring the whole chunk as though it were an unknown chunk
type. (This recommendation assumes that the chunk's CRC has been
verified. In decoders that do not check CRCs, it is safer to
treat any unexpected value as indicating a corrupted
datastream.)</p>

<p>Standard PNG images shall be compressed with compression
method 0. The compression method field of the <a href="#11IHDR"><span class=
"chunk">IHDR</span></a> chunk is
provided for possible future standardization or proprietary
variants. Decoders shall check this byte and report an error if
it holds an unrecognized code. See clause&#160;10: <a href=
"#10Compression"><span class="xref">Compression</span></a> for
details.</p>

<h2><a name="13Security-considerations">13.4 Security
considerations</a></h2>

<p>A PNG datastream is composed of a collection of explicitly
typed chunks. Chunks whose contents are defined by the
specification could actually contain anything, including
malicious code. But there is no known risk that such malicious
code could be executed on the recipient's computer as a result of
decoding the PNG image.</p>

<p>The possible security risks associated with future chunk types
cannot be specified at this time. Security issues will be
considered by the Registration Authority when evaluating chunks
proposed for registration as public chunks. There is no
additional security risk associated with unknown or unimplemented
chunk types, because such chunks will be ignored, or at most be
copied into another PNG datastream.</p>

<p>The <a href="#11iTXt"><span class="chunk">iTXt</span></a>, <a
href="#11tEXt"><span class="chunk">tEXt</span></a>, and <a href=
"#11zTXt"><span class="chunk">zTXt</span></a> chunks contain keywords
and data
that are meant to be displayed as plain text. The <a href=
"#11iCCP"><span class="chunk">iCCP</span></a> and <a href=
"#11sPLT"><span class="chunk">sPLT</span></a> chunks contain
keywords that are meant to be displayed as plain text. It is
possible that if the decoder displays such text without filtering
out control characters, especially the ESC (escape) character,
certain systems or terminals could behave in undesirable and
insecure ways. It is recommended that decoders filter out control
characters to avoid this risk; see 13.5.3: <a href=
"#13Text-chunk-processing"><span class="xref">Text chunk
processing</span></a>.</p>

<p>Every chunk begins with a length field, which makes it easier
to write decoders that are invulnerable to fraudulent chunks that
attempt to overflow buffers. The CRC at the end of every chunk
provides a robust defence against accidentally corrupted data.
The PNG signature bytes provide early detection of common file
transmission errors.</p>

<p>A decoder that fails to check CRCs could be subject to data
corruption. The only likely consequence of such corruption is
incorrectly displayed pixels within the image. Worse things might
happen if the CRC of the <a href="#11IHDR"><span class=
"chunk">IHDR</span></a> chunk is not checked and the width or
height fields are corrupted. See 13.3: <a href=
"#13Error-checking"><span class="xref">Error
checking</span></a>.</p>

<p>A poorly written decoder might be subject to buffer overflow,
because chunks can be extremely large, up to 2<sup>31</sup>-1
bytes long. But properly written decoders will handle large
chunks without difficulty.</p>

<h2><a name="13Chunking">13.5 Chunking</a></h2>

<p>Decoders shall recognize chunk types by a simple four-byte
literal comparison; it is incorrect to perform case conversion on
chunk types. A decoder encountering an unknown chunk in which the
ancillary bit is 1 may safely ignore the chunk and proceed to
display the image. A decoder trying to extract the image, upon
encountering an unknown chunk in which the ancillary bit is 0,
indicating a critical chunk, shall indicate to the user that the
image contains information it cannot safely interpret.</p>

<p>(Decoders should not flag an error if the reserved bit is set
to 1, however, as some future version of the PNG specification
could define a meaning for this bit. It is sufficient to treat a
chunk with this bit set in the same way as any other unknown
chunk type.)</p>

<h2><a name="13Pixel-dimensions">13.6 Pixel dimensions</a></h2>

<p>Non-square pixels can be represented (see 11.3.5.3: <a href=
"#11pHYs"><span class="xref"><span class="chunk">pHYs</span>
Physical pixel dimensions</span></a>), but viewers are not
required to account for them; a viewer can present any PNG
datastream as though its pixels are square.</p>

<p>Where the pixel aspect ratio of the display differs from the
aspect ratio of the physical pixel dimensions defined in the PNG
datastream, viewers are strongly encouraged to rescale images for
proper display.</p>

<p>When the <a href="#11pHYs"><span class="xref"><span class=
"chunk">pHYs</span></span></a> chunk has a unit specifier of 0
(unit is unknown), the behaviour of a decoder may depend on the
ratio of the two pixels-per-unit values, but should not depend on
their magnitudes. For example, a <a href="#11pHYs"><span class=
"xref"><span class="chunk">pHYs</span></span></a> chunk
containing <tt>(ppuX, ppuY, unit) = (2, 1, 0)</tt> is equivalent
to one containing <tt>(1000, 500, 0)</tt>; both are equally valid
indications that the image pixels are twice as tall as they are
wide.</p>

<p>One reasonable way for viewers to handle a difference between
the pixel aspect ratios of the image and the display is to expand
the image either horizontally or vertically, but not both. The
scale factors could be obtained using the following
floating-point calculations:</p>

<pre>
<tt>image_ratio = pHYs_ppuY / pHYs_ppuX
display_ratio = display_ppuY / display_ppuX
scale_factor_X = max(1.0, image_ratio/display_ratio)
scale_factor_Y = max(1.0, display_ratio/image_ratio)</tt>
</pre>

<p>Because other methods such as maintaining the image area are
also reasonable, and because ignoring the <a href="#11pHYs"><span
class="xref"><span class="chunk">pHYs</span></span></a> chunk is
permissible, authors should not assume that all viewing
applications will use this scaling method.</p>

<p>As well as making corrections for pixel aspect ratio, a viewer
may have reasons to perform additional scaling both horizontally
and vertically. For example, a viewer might want to shrink an
image that is too large to fit on the display, or to expand
images sent to a high-resolution printer so that they appear the
same size as they did on the display.</p>

<h2><a name="13Text-chunk-processing">13.7 Text chunk
processing</a></h2>

<p>If practical, PNG decoders should have a way to display to the
user all the <a href="#11iTXt"><span class=
"chunk">iTXt</span></a>, <a href="#11tEXt"><span class=
"chunk">tEXt</span></a>, and <a href="#11zTXt"><span class=
"chunk">zTXt</span></a> chunks found in the datastream. Even if
the decoder does not recognize a particular text keyword, the
user might be able to understand it.</p>

<p>When processing <a href="#11tEXt"><span class=
"chunk">tEXt</span></a> and <a href="#11zTXt"><span class=
"chunk">zTXt</span></a> chunks, decoders could encounter
characters other than those permitted. Some can be safely
displayed (e.g., TAB, FF, and CR, decimal 9, 12, and 13,
respectively), but others, especially the ESC character (decimal
27), could pose a security hazard (because unexpected actions may
be taken by display hardware or software). Decoders should not
attempt to directly display any non-Latin-1 characters (except
for newline and perhaps TAB, FF, CR) encountered in a <a href=
"#11tEXt"><span class="chunk">tEXt</span></a> or <a href=
"#11zTXt"><span class="chunk">zTXt</span></a> chunk. Instead,
they should be ignored or displayed in a visible notation such as
"<tt>\</tt>nnn". See 13.4: <a href=
"#13Security-considerations"><span class="xref">Security
considerations</span></a>.</p>

<p>Even though encoders are recommended to represent newlines as
linefeed (decimal 10), it is recommended that decoders not rely
on this; it is best to recognize all the common newline
combinations (CR, LF, and CR-LF) and display each as a single
newline. TAB can be expanded to the proper number of spaces
needed to arrive at a column multiple of 8.</p>

<p>Decoders running on systems with non-Latin-1 character set
encoding should provide character code remapping so that Latin-1
characters are displayed correctly. Some systems may not provide
all the characters defined in Latin-1. Mapping unavailable
characters to a visible notation such as "<tt>\</tt>nnn" is a
good fallback. Character codes 127-255 should be displayed only
if they are printable characters on the decoding system. Some
systems may interpret such codes as control characters; for
security, decoders running on such systems should not display
such characters literally.</p>

<p>Decoders should be prepared to display text chunks that
contain any number of printing characters between newline
characters, even though it is recommended that encoders avoid
creating lines in excess of 79 characters.</p>

<h2><a name="13Decompression">13.8 Decompression</a></h2>

<p>The compression technique used in this International Standard
does not require the entire compressed datastream to be available
before decompression can start. Display can therefore commence
before the entire decompressed datastream is available. It is
extremely unlikely that any general purpose compression methods
in future versions of this International Standard will not have
this property.</p>

<p>It is important to emphasize that <a href="#11IDAT"><span
class="chunk">IDAT</span></a> chunk boundaries have no semantic
significance and can occur at any point in the compressed
datastream. There is no required correlation between the
structure of the image data (for example, scanline boundaries) and
deflate block boundaries or <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunk boundaries. The complete image data
is represented by a single zlib datastream that is stored in some
number of <a href="#11IDAT"><span class="chunk">IDAT</span></a>
chunks; a decoder that assumes any more than this is incorrect.
Some encoder implementations may emit datastreams in which some
of these structures are indeed related, but decoders cannot rely
on this.</p>

<h2><a name="13Filtering">13.9 Filtering</a></h2>

<p>To reverse the effect of a filter, the decoder may need
to use the decoded values of the prior pixel on the same line,
the pixel immediately above the current pixel on the prior line,
and the pixel just to the left of the pixel above. This implies
that at least one scanline's worth of image data needs to be
stored by the decoder at all times. Even though some filter types
do not refer to the prior scanline, the decoder will always need
to store each scanline as it is decoded, since the next scanline
might use a filter type that refers to it.</p>

<h2><a name="13Progressive-display">13.10 Interlacing and
progressive display</a></h2>

<p>Decoders are required to be able to read interlaced images. If
the reference image contains fewer than five columns or fewer
than five rows, some passes will be empty. Encoders and decoders
shall handle this case correctly. In particular, filter type
bytes are associated only with nonempty scanlines; no filter type
bytes are present in an empty reduced image.</p>

<p>When receiving images over slow transmission links, viewers
can improve perceived performance by displaying interlaced images
progressively. This means that as each reduced image is received,
an approximation to the complete image is displayed based on the
data received so far. One simple yet pleasing effect can be
obtained by expanding each received pixel to fill a rectangle
covering the yet-to-be-transmitted pixel positions below and to
the right of the received pixel. This process can be described by
the following ISO C code <a href="#2-ISO-9899"><span class=
"NormRef">[ISO-9899]</span></a>:</p>

<pre>
/*
    variables declared and initialized elsewhere in the code:
        height, width
    functions or macros defined elsewhere in the code:
        visit(), min()
 */

int starting_row[7]  = { 0, 0, 4, 0, 2, 0, 1 };
int starting_col[7]  = { 0, 4, 0, 2, 0, 1, 0 };
int row_increment[7] = { 8, 8, 8, 4, 4, 2, 2 };
int col_increment[7] = { 8, 8, 4, 4, 2, 2, 1 };
int block_height[7]  = { 8, 8, 4, 4, 2, 2, 1 };
int block_width[7]   = { 8, 4, 4, 2, 2, 1, 1 };

int pass;
long row, col;
   
pass = 0;
while (pass &lt; 7)
{
    row = starting_row[pass];
    while (row &lt; height)
    {
        col = starting_col[pass];
        while (col &lt; width)
        {
            visit(row, col,
                  min(block_height[pass], height - row),
                  min(block_width[pass], width - col));
            col = col + col_increment[pass];
        }
        row = row + row_increment[pass];
    }
    pass = pass + 1;
}
</pre>

<p>The function <tt>visit(row,column,height,width)</tt> obtains
the next transmitted pixel and paints a rectangle of the
specified height and width, whose upper-left corner is at the
specified row and column, using the colour indicated by the
pixel. Note that row and column are measured from 0,0 at the
upper left corner.</p>

<p>If the viewer is merging the received image with a background
image, it may be more convenient just to paint the received pixel
positions (the <tt>visit()</tt> function sets only the pixel at the
specified row and column, not the whole rectangle). This produces
a "fade-in" effect as the new image gradually replaces the old.
An advantage of this approach is that proper alpha or
transparency processing can be done as each pixel is replaced.
Painting a rectangle as described above will overwrite
background-image pixels that may be needed later, if the pixels
eventually received for those positions turn out to be wholly or
partially transparent. This is a problem only if the background
image is not stored anywhere offscreen.</p>

<h2><a name="13Truecolour-image-handling">13.11 Truecolour image
handling</a></h2>

<p>To achieve PNG's goal of universal interchangeability,
decoders shall accept all types of PNG image: indexed-colour,
truecolour, and greyscale. Viewers running on indexed-colour
display hardware need to be able to reduce truecolour images to
indexed-colour for viewing. This process is called "colour
quantization".</p>

<p>A simple, fast method for colour quantization is to reduce the
image to a fixed palette. Palettes with uniform colour spacing
("colour cubes") are usually used to minimize the per-pixel
computation. For photograph-like images, dithering is recommended
to avoid ugly contours in what should be smooth gradients;
however, dithering introduces graininess that can be
objectionable.</p>

<p>The quality of rendering can be improved substantially by
using a palette chosen specifically for the image, since a colour
cube usually has numerous entries that are unused in any
particular image. This approach requires more work, first in
choosing the palette, and second in mapping individual pixels to
the closest available colour. PNG allows the encoder to supply
suggested palettes, but not all encoders will do so, and the
suggested palettes may be unsuitable in any case (they may have
too many or too few colours). Therefore, high-quality viewers
will need to have a palette selection routine at hand. A large
lookup table is usually the most feasible way of mapping
individual pixels to palette entries with adequate speed.</p>

<p>Numerous implementations of colour quantization are available.
The PNG sample implementation, libpng (<a href=
"http://www.libpng.org/pub/png/libpng.html"><code>http://www.libpng.org/pub/png/libpng.html</code></a>),
includes code for the purpose.</p>

<h2><a name="13Sample-depth-rescaling">13.12 Sample depth
rescaling</a></h2>

<p>Decoders may wish to scale PNG data to a lesser sample depth
(data precision) for display. For example, 16-bit data will need
to be reduced to 8-bit depth for use on most present-day display
hardware. Reduction of 8-bit data to 5-bit depth is also
common.</p>

<p>The most accurate scaling is achieved by the linear
equation</p>

<p><tt>output = floor((input * MAXOUTSAMPLE / MAXINSAMPLE) +
0.5)</tt></p>

<p>where</p>

<p><tt>MAXINSAMPLE = (2<sup>sampledepth</sup>)-1</tt><br class="xhtml" />
 <tt>MAXOUTSAMPLE = (2<sup>desired_sampledepth</sup>)-1</tt></p>

<p>A slightly less accurate conversion is achieved by simply
shifting right by <tt>(sampledepth - desired_sampledepth)</tt>
places. For example, to reduce 16-bit samples to 8-bit, the
low-order byte can be discarded. In many situations the shift
method is sufficiently accurate for display purposes, and it is
certainly much faster. (But if gamma correction is being done,
sample rescaling can be merged into the gamma correction lookup
table, as is illustrated in 13.13: <a href=
"#13Decoder-gamma-handling"><span class="xref">Decoder gamma
handling</span></a>.)</p>

<p>If the decoder needs to scale samples up (for example, if the
frame buffer has a greater sample depth than the PNG image), it
should use linear scaling or left-bit-replication as described in
12.5: <a href="#12Sample-depth-scaling"><span class="xref">Sample
depth scaling</span></a>.</p>

<p>When an <a href="#11sBIT"><span class="chunk">sBIT</span></a>
chunk is present, the reference image data can be recovered by
shifting right to the sample depth specified by <a href=
"#11sBIT"><span class="chunk">sBIT</span></a>. Note that linear
scaling will not necessarily reproduce the original data, because
the encoder is not required to have used linear scaling to scale
the data up. However, the encoder is required to have used a
method that preserves the high-order bits, so shifting always
works. This is the only case in which shifting might be said to
be more accurate than linear scaling. A decoder need not pay
attention to the <a href="#11sBIT"><span class=
"chunk">sBIT</span></a> chunk; the stored image is a valid PNG
datastream of the sample depth indicated by the <a href=
"#11IHDR"><span class="chunk">IHDR</span></a> chunk; however,
using <a href="#11sBIT"><span class="chunk">sBIT</span></a> to
recover the original samples before scaling them to suit the
display often yields a more accurate display than ignoring <a
href="#11sBIT"><span class="chunk">sBIT</span></a>.</p>

<p>When comparing pixel values to <a href="#11tRNS"><span class=
"chunk">tRNS</span></a> chunk values to detect transparent
pixels, the comparison shall be done exactly. Therefore,
transparent pixel detection shall be done before reducing sample
precision.</p>

<h2><a name="13Decoder-gamma-handling">13.13 Decoder gamma
handling</a></h2>

<p>See Annex C: <a href="#C-GammaAppendix"><span class=
"xref">Gamma and chromaticity</span></a> for a brief introduction
to gamma issues.</p>

<p>Viewers capable of full colour management <a href=
"#G-ICC"><span class="bibref">[ICC]</span></a> will perform more
sophisticated calculations than those described here.</p>

<p>For an image display program to produce correct tone
reproduction, it is necessary to take into account the
relationship between samples and display output, and the transfer
function of the display system. This can be done by
calculating:</p>

<p><tt>sample = integer_sample / (2<sup>sampledepth</sup> -
1.0)<br class="xhtml" />
 display_output = sample<sup>1.0/gamma</sup><br class="xhtml" />
 display_input = inverse_display_transfer(display_output)<br class="xhtml" />
 framebuf_sample = floor((display_input *
MAX_FRAMEBUF_SAMPLE)+0.5)</tt></p>

<p>where <tt>integer_sample</tt> is the sample value from the
datastream, <tt>framebuf_sample</tt> is the value to write into
the frame buffer, and <tt>MAX_FRAMEBUF_SAMPLE</tt> is the maximum
value of a frame buffer sample (255 for 8-bit, 31 for 5-bit,
etc). The first line converts an integer sample into a normalized
floating point value (in the range 0.0 to 1.0), the second
converts to a value proportional to the desired display output
intensity, the third accounts for the display system's transfer
function, and the fourth converts to an integer frame buffer
sample. Zero raised to any positive power is zero.</p>

<p>A step could be inserted between the second and third to
adjust <tt>display_output</tt> to account for the difference
between the actual viewing conditions and the reference viewing
conditions. However, this adjustment requires accounting for
veiling glare, black mapping, and colour appearance models, none
of which can be well approximated by power functions. Such
calculations are not described here. If viewing conditions are
ignored, the error will usually be small.</p>

<p>The display transfer function can typically be approximated by
a power function with exponent <tt>display_exponent</tt>, in
which case the second and third lines can be merged into:</p>

<p><tt>display_input = sample<sup>1.0/(gamma *
display_exponent)</sup> =
sample<sup>decoding_exponent</sup></tt></p>

<p>so as to perform only one power calculation. For colour
images, the entire calculation is performed separately for R, G,
and B values.</p>

<p>The value of gamma can be taken directly from the <a href=
"#11gAMA"><span class="chunk">gAMA</span></a> chunk.
Alternatively, an application may wish to allow the user to
adjust the appearance of the displayed image by influencing the
value of gamma. For example, the user could manually set a
parameter <tt>user_exponent</tt> which defaults to 1.0, and the
application could set:</p>

<pre>
<tt>gamma = gamma_from_file / user_exponent
decoding_exponent = 1.0 / (gamma * display_exponent)
   = user_exponent / (gamma_from_file * display_exponent)</tt>
</pre>

<p>The user would set <tt>user_exponent</tt> greater than 1 to
darken the mid-level tones, or less than 1 to lighten them.</p>

<p>A <a href=
"#11gAMA"><span class="chunk">gAMA</span></a> chunk containing zero is
meaningless but could appear by mistake.
Decoders should ignore it,
and editors may discard it and issue a warning to the user.</p>

<p>It is <strong>not</strong> necessary to perform a transcendental
mathematical computation for every pixel. Instead, a lookup table
can be computed that gives the correct output value for every
possible sample value. This requires only 256 calculations per
image (for 8-bit accuracy), not one or three calculations per
pixel. For an indexed-colour image, a one-time correction of the
palette is sufficient, unless the image uses transparency and is
being displayed against a nonuniform background.</p>

<p>If floating-point calculations are not possible, gamma
correction tables can be computed using integer arithmetic and a
precomputed table of logarithms. Example code appears in <a href=
"#G-PNG-EXTENSIONS"><span class=
"bibref">[PNG-EXTENSIONS]</span></a>.</p>

<p>When the incoming image has unknown gamma (<a href=
"#11gAMA"><span class="chunk">gAMA</span></a>, <a href=
"#11sRGB"><span class="chunk">sRGB</span></a>, and <a href=
"#11iCCP"><span class="chunk">iCCP</span></a> all absent), choose
a likely default gamma value, but allow the user to select a new
one if the result proves too dark or too light. The default gamma
may depend on other knowledge about the image, for example
whether it came from the Internet or from the local system.</p>

<p>In practice, it is often difficult to determine what value of
display exponent should be used. In systems with no built-in
gamma correction, the display exponent is determined entirely by
the CRT. A display exponent of 2.2 should be used unless detailed
calibration measurements are available for the particular CRT
used.</p>

<p>Many modern frame buffers have lookup tables that are used to
perform gamma correction, and on these systems the display
exponent value should be the exponent of the lookup table and CRT
combined. It may not be possible to find out what the lookup
table contains from within the viewer application, in which case
it may be necessary to ask the user to supply the display
system's exponent value. Unfortunately, different manufacturers
use different ways of specifying what should go into the lookup
table, so interpretation of the system "gamma" value is
system-dependent.</p>

<p>The response of real displays is actually more complex than
can be described by a single number (the display exponent). If
actual measurements of the monitor's light output as a function
of voltage input are available, the third and fourth lines of the
computation above can be replaced by a lookup in these
measurements, to find the actual frame buffer value that most
nearly gives the desired brightness.</p>

<h2><a name="13Decoder-colour-handling">13.14 Decoder colour
handling</a></h2>

<p>See Annex C: <a href="#C-GammaAppendix"><span class=
"xref">Gamma and chromaticity</span></a> for references to colour
issues.</p>

<p>In many cases, the image data in PNG datastreams will be
treated as device-dependent RGB values and displayed without
modification (except for appropriate gamma correction). This
provides the fastest display of PNG images. But unless the viewer
uses exactly the same display hardware as that used by the author
of the original image, the colours will not be exactly the same
as those seen by the original author, particularly for darker or
near-neutral colours. The <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunk provides information that allows
closer colour matching than that provided by gamma correction
alone.</p>

<p>The <a href="#11cHRM"><span class="chunk">cHRM</span></a> data
can be used to transform the image data from RGB to XYZ and
thence into a perceptually linear colour space such as CIE LAB.
The colours can be partitioned to generate an optimal palette,
because the geometric distance between two colours in CIE LAB is
strongly related to how different those colours appear (unlike,
for example, RGB or XYZ spaces). The resulting palette of
colours, once transformed back into RGB colour space, could be
used for display or written into a <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk.</p>

<p>Decoders that are part of image processing applications might
also transform image data into CIE LAB space for analysis.</p>

<p>In applications where colour fidelity is critical, such as
product design, scientific visualization, medicine, architecture,
or advertising, PNG decoders can transform the image data from
source RGB to the display RGB space of the monitor used to view
the image. This involves calculating the matrix to go from source
RGB to XYZ and the matrix to go from XYZ to display RGB, then
combining them to produce the overall transformation. The PNG
decoder is responsible for implementing gamut mapping.</p>

<p>Decoders running on platforms that have a Colour Management
System (CMS) can pass the image data, <a href="#11gAMA"><span
class="chunk">gAMA</span></a>, and <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> values to the CMS for display or further
processing.</p>

<p>PNG decoders that provide colour printing facilities can use
the facilities in Level 2 PostScript to specify image data in
calibrated RGB space or in a device-independent colour space such
as XYZ. This will provide better colour fidelity than a simple
RGB to CMYK conversion. The PostScript Language Reference manual
<a href="#G-POSTSCRIPT"><span class=
"bibref">[POSTSCRIPT]</span></a> gives examples. Such decoders
are responsible for implementing gamut mapping between source RGB
(specified in the <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> chunk) and the target printer. The
PostScript interpreter is then responsible for producing the
required colours.</p>

<p>PNG decoders can use the <a href="#11cHRM"><span class=
"chunk">cHRM</span></a> data to calculate an accurate greyscale
representation of a colour image. Conversion from RGB to grey is
simply a case of calculating the Y (luminance) component of XYZ,
which is a weighted sum of R, G, and B values. The weights depend
upon the monitor type, i.e. the values in the <a href=
"#11cHRM"><span class="chunk">cHRM</span></a> chunk. PNG decoders
may wish to do this for PNG datastreams with no <a href=
"#11cHRM"><span class="chunk">cHRM</span></a> chunk. In this
case, a reasonable default would be the CCIR 709 primaries <a
href="#G-ITU-R-BT709"><span class=
"bibref">[ITU-R-BT709]</span></a>. The original NTSC primaries
should <strong>not</strong> be used unless the PNG image really
was colour-balanced for such a monitor.</p>

<h2><a name="13Background-colour">13.15 Background
colour</a></h2>

<p>The background colour given by the <a href="#11bKGD"><span
class="chunk">bKGD</span></a> chunk will typically be used to
fill unused screen space around the image, as well as any
transparent pixels within the image. (Thus, <a href=
"#11bKGD"><span class="chunk">bKGD</span></a> is valid and useful
even when the image does not use transparency.) If no <a href=
"#11bKGD"><span class="chunk">bKGD</span></a> chunk is present,
the viewer will need to decide upon a suitable background colour.
When no other information is available, a medium grey such as 153
in the 8-bit sRGB colour space would be a reasonable choice.
Transparent black or white text and dark drop shadows, which are
common, would all be legible against this background.</p>

<p>Viewers that have a specific background against which to
present the image (such as web browsers) should ignore the <a
href="#11bKGD"><span class="chunk">bKGD</span></a> chunk, in
effect overriding <a href="#11bKGD"><span class=
"chunk">bKGD</span></a> with their preferred background colour or
background image.</p>

<p>The background colour given by the <a href="#11bKGD"><span
class="chunk">bKGD</span></a> chunk is not to be considered
transparent, even if it happens to match the colour given by the
<a href="#11tRNS"><span class="chunk">tRNS</span></a> chunk (or,
in the case of an indexed-colour image, refers to a palette index
that is marked as transparent by the <a href="#11tRNS"><span
class="chunk">tRNS</span></a> chunk). Otherwise one would have to
imagine something "behind the background" to composite against.
The background colour is either used as background or ignored; it
is not an intermediate layer between the PNG image and some other
background.</p>

<p>Indeed, it will be common that the <a href="#11bKGD"><span
class="chunk">bKGD</span></a> and <a href="#11tRNS"><span class=
"chunk">tRNS</span></a> chunks specify the same colour, since
then a decoder that does not implement transparency processing
will give the intended display, at least when no
partially-transparent pixels are present.</p>

<h2><a name="13Alpha-channel-processing">13.16 Alpha channel
processing</a></h2>

<p>The alpha channel can be used to composite a foreground image
against a background image. The PNG datastream defines the
foreground image and the transparency mask, but not the
background image. PNG decoders are <strong>not</strong> required to
support this most general case. It is expected that most will be
able to support compositing against a single background
colour.</p>

<p>The equation for computing a composited sample value is:</p>

<pre>
output = alpha * foreground + (1-alpha) * background
</pre>

<p>where alpha and the input and output sample values are
expressed as fractions in the range 0 to 1. This computation
should be performed with intensity samples (not gamma-encoded
samples). For colour images, the computation is done separately
for R, G, and B samples.</p>

<p>The following code illustrates the general case of compositing
a foreground image against a background image. It assumes that
the original pixel data are available for the background image,
and that output is to a frame buffer for display. Other variants
are possible; see the comments below the code. The code allows
the sample depths and gamma values of foreground image and
background image all to be different and not necessarily suited
to the display system. In practice no assumptions about equality
should be made without first checking.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p>This code is ISO C <a href="#2-ISO-9899"><span class=
"NormRef">[ISO-9899]</span></a>, with line numbers added for
reference in the comments below.</p>

<pre>
   01  int foreground[4];  /* image pixel: R, G, B, A */
   02  int background[3];  /* background pixel: R, G, B */
   03  int fbpix[3];       /* frame buffer pixel */
   04  int fg_maxsample;   /* foreground max sample */
   05  int bg_maxsample;   /* background max sample */
   06  int fb_maxsample;   /* frame buffer max sample */
   07  int ialpha;
   08  float alpha, compalpha;
   09  float gamfg, linfg, gambg, linbg, comppix, gcvideo;
   
       /* Get max sample values in data and frame buffer */
   10  fg_maxsample = (1 &lt;&lt; fg_sample_depth) - 1;
   11  bg_maxsample = (1 &lt;&lt; bg_sample_depth) - 1;
   12  fb_maxsample = (1 &lt;&lt; frame_buffer_sample_depth) - 1;
       /*
        * Get integer version of alpha.
        * Check for opaque and transparent special cases;
        * no compositing needed if so.
        *
        * We show the whole gamma decode/correct process in
        * floating point, but it would more likely be done
        * with lookup tables.
        */
   13  ialpha = foreground[3];
   
   14  if (ialpha == 0) {
           /*
            * Foreground image is transparent here.
            * If the background image is already in the frame
            * buffer, there is nothing to do.
            */
   15      ;
   16  } else if (ialpha == fg_maxsample) {
           /*
            * Copy foreground pixel to frame buffer.
            */
   17      for (i = 0; i &lt; 3; i++) {
   18          gamfg = (float) foreground[i] / fg_maxsample;
   19          linfg = pow(gamfg, 1.0 / fg_gamma);
   20          comppix = linfg;
   21          gcvideo = pow(comppix, 1.0 / display_exponent);
   22          fbpix[i] = (int) (gcvideo * fb_maxsample + 0.5);
   23      }
   24  } else {
           /*
            * Compositing is necessary.
            * Get floating-point alpha and its complement.
            * Note: alpha is always linear; gamma does not
            * affect it.
            */
   25      alpha = (float) ialpha / fg_maxsample;
   26      compalpha = 1.0 - alpha;
   
   27      for (i = 0; i &lt; 3; i++) {
               /*
                * Convert foreground and background to floating
                * point, then undo gamma encoding.
                */
   28          gamfg = (float) foreground[i] / fg_maxsample;
   29          linfg = pow(gamfg, 1.0 / fg_gamma);
   30          gambg = (float) background[i] / bg_maxsample;
</pre>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<pre>
   31          linbg = pow(gambg, 1.0 / bg_gamma);
               /* 
                * Composite.
                */
   32          comppix = linfg * alpha + linbg * compalpha;
               /*
                * Gamma correct for display.
                * Convert to integer frame buffer pixel.
                */
   33          gcvideo = pow(comppix, 1.0 / display_exponent);
   34          fbpix[i] = (int) (gcvideo * fb_maxsample + 0.5);
   35      }
   36  }
</pre>

<p>Variations:</p>

<!-- <ol start="1"> --><ol>
<li>If output is to another PNG datastream instead of a frame
buffer, lines 21, 22, 33, and 34 should be changed along the
following lines 

<pre>
   /*
    * Gamma encode for storage in output datastream.
    * Convert to integer sample value.
    */
   gamout = pow(comppix, outfile_gamma);
   outpix[i] = (int) (gamout * out_maxsample + 0.5);
</pre>

Also, it becomes necessary to process background pixels when
alpha is zero, rather than just skipping pixels. Thus, line 15
will need to be replaced by copies of lines 17-23, but processing
background instead of foreground pixel values.</li>

<li>If the sample depths of the output file, foreground file, and
background file are all the same, and the three gamma values also
match, then the no-compositing code in lines 14-23 reduces to
copying pixel values from the input file to the output file if
alpha is one, or copying pixel values from background to output
file if alpha is zero. Since alpha is typically either zero or
one for the vast majority of pixels in an image, this is a
significant saving. No gamma computations are needed for most
pixels.</li>

<li>When the sample depths and gamma values all match, it may
appear attractive to skip the gamma decoding and encoding (lines
28-31, 33-34) and just perform line 32 using gamma-encoded sample
values. Although this does not have too bad an effect on image
quality, the time savings are small if alpha values of zero and
one are treated as special cases as recommended here.</li>

<li>If the original pixel values of the background image are no
longer available, only processed frame buffer pixels left by
display of the background image, then lines 30 and 31 need to
extract intensity from the frame buffer pixel values using code
such as 

<pre>
   /*
    * Convert frame buffer value into intensity sample.
    */
   gcvideo = (float) fbpix[i] / fb_maxsample;
   linbg = pow(gcvideo, display_exponent);
</pre>

However, some roundoff error can result, so it is better to have
the original background pixels available if at all possible.</li>

<li>Note that lines 18-22 are performing exactly the same gamma
computation that is done when no alpha channel is present. If the
no-alpha case is handled with a lookup table, the same lookup
table can be used here. Lines 28-31 and 33-34 can also be done
with (different) lookup tables.</li>

<li>Integer arithmetic can be used instead of floating point,
providing care is taken to maintain sufficient precision
throughout.</li>
</ol>

<p class="Note">NOTE In floating point, no overflow or underflow
checks are needed, because the input sample values are guaranteed
to be between 0 and 1, and compositing always yields a result
that is in between the input values (inclusive). With integer
arithmetic, some roundoff-error analysis might be needed to
guarantee no overflow or underflow.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<p>When displaying a PNG image with full alpha channel, it is
important to be able to composite the image against some
background, even if it is only black. Ignoring the alpha channel
will cause PNG images that have been converted from an
associated-alpha representation to look wrong. (Of course, if the
alpha channel is a separate transparency mask, then ignoring
alpha is a useful option: it allows the hidden parts of the image
to be recovered.)</p>

<p>Even if the decoder does not implement true compositing logic,
it is simple to deal with images that contain only zero and one
alpha values. (This is implicitly true for greyscale and
truecolour PNG datastreams that use a <a href="#11tRNS"><span
class="chunk">tRNS</span></a> chunk; for indexed-colour PNG
datastreams it is easy to check whether the <a href=
"#11tRNS"><span class="chunk">tRNS</span></a> chunk contains any
values other than 0 and 255.) In this simple case, transparent
pixels are replaced by the background colour, while others are
unchanged.</p>

<p>If a decoder contains only this much transparency capability,
it should deal with a full alpha channel by treating all nonzero
alpha values as fully opaque or by dithering. Neither approach
will yield very good results for images converted from
associated-alpha formats, but this is preferable to doing
nothing. Dithering full alpha to binary alpha is very much like
dithering greyscale to black-and-white, except that all fully
transparent and fully opaque pixels should be left unchanged by
the dither.</p>

<h2><a name="13Histogram-and-suggested-palette-usage">13.17
Histogram and suggested palette usage</a></h2>

<p>For viewers running on indexed-colour hardware attempting to
display a truecolour image, or an indexed-colour image whose
palette is too large for the frame buffer, the encoder may have
provided one or more suggested palettes in <a href=
"#11sPLT"><span class="chunk">sPLT</span></a> chunks. If one of
these is found to be suitable, based on size and perhaps name,
the PNG decoder can use that palette. Suggested palettes with a
sample depth different from what the decoder needs can be
converted using sample depth rescaling (see 13.12: <a href=
"#13Sample-depth-rescaling"><span class="xref">Sample depth
rescaling</span></a>).</p>

<p>When the background is a solid colour, the viewer should
composite the image and the suggested palette against that
colour, then quantize the resulting image to the resulting RGB
palette. When the image uses transparency and the background is
not a solid colour, no suggested palette is likely to be
useful.</p>

<p>For truecolour images, a suggested palette might also be
provided in a <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk. If the image has a <a href=
"#11tRNS"><span class="chunk">tRNS</span></a> chunk and the
background is a solid colour, the viewer will need to adapt the
suggested palette for use with its desired background colour. To
do this, the palette entry closest to the <a href="#11tRNS"><span
class="chunk">tRNS</span></a> colour should be replaced with the
desired background colour; or alternatively a palette entry for
the background colour can be added, if the viewer can handle more
colours than there are <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> entries.</p>

<p>For images of colour type 6 (truecolour with alpha), any <a
href="#11PLTE"><span class="chunk">PLTE</span></a> chunk should
have been designed for display of the image against a uniform
background of the colour specified by the <a href="#11bKGD"><span
class="chunk">bKGD</span></a> chunk. Viewers should probably
ignore the palette if they intend to use a different background,
or if the <a href="#11bKGD"><span class="chunk">bKGD</span></a>
chunk is missing. Viewers can use a suggested palette for display
against a different background than it was intended for, but the
results may not be very good.</p>

<p>If the viewer presents a transparent truecolour image against
a background that is more complex than a uniform colour, it is
unlikely that the suggested palette will be optimal for the
composite image. In this case it is best to perform a truecolour
compositing step on the truecolour PNG image and background
image, then colour-quantize the resulting image.</p>

<p>In truecolour PNG datastreams, if both <a href="#11PLTE"><span
class="chunk">PLTE</span></a> and <a href="#11sPLT"><span class=
"chunk">sPLT</span></a> chunks appear, the PNG decoder may choose
from among the palettes suggested by both, bearing in mind the
different transparency semantics described above.</p>

<p>The frequencies in the <a href="#11sPLT"><span class=
"chunk">sPLT</span></a> and <a href="#11hIST"><span class=
"chunk">hIST</span></a> chunks are useful when the viewer cannot
provide as many colours as are used in the palette in the PNG
datastream. If the viewer has a shortfall of only a few colours,
it is usually adequate to drop the least-used colours from the
palette. To reduce the number of colours substantially, it is
best to choose entirely new representative colours, rather than
trying to use a subset of the existing palette. This amounts to
performing a new colour quantization step; however, the existing
palette and histogram can be used as the input data, thus
avoiding a scan of the image data in the <a href="#11IDAT"><span
class="chunk">IDAT</span></a> chunks.</p>

<p>If no suggested palette is provided, a decoder can develop its
own, at the cost of an extra pass over the image data in the <a
href="#11IDAT"><span class="chunk">IDAT</span></a> chunks.
Alternatively, a default palette (probably a colour cube) can be
used.</p>

<p>See also 12.6: <a href="#12Suggested-palettes"><span class=
"xref">Suggested palettes</span></a>.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="14EditorsExt">14 Editors and extensions</a></h1>

<h2><a name="14Additional-chunk-types">14.1 Additional chunk
types</a></h2>

<p>The provisions of this International Standard may be extended
by adding new chunk types, which may be either private or public.
Applications can use private chunk types to carry data that is
not of interest to other people's applications.</p>

<p>Decoders shall be prepared to encounter unrecognized public or
private chunk types. The chunk naming conventions (see 5.4:
<a href="#5Chunk-naming-conventions"><span class="xref">Chunk
naming conventions</span></a>) enable critical/ancillary,
public/private, and safe/unsafe to copy chunks to be
distinguished.</p>

<p>Additional public PNG chunk types are defined in the document
Register of PNG Public Chunks and Keywords <a href=
"#G-PNG-EXTENSIONS"><span class=
"bibref">[PNG-REGISTER]</span></a>. Chunks described there are
expected to be less widely supported than those defined in this
International Standard. However, application authors are
encouraged to use those chunk types whenever appropriate for
their applications. Additional chunk types can be proposed for
inclusion in that list by contacting the PNG Registration
Authority (see 4.9: <a href="#4Concepts.Registration"><span
class="xref">Extension and registration</span></a>).</p>

<p>New public chunks will be registered only if they are of use
to others and do not violate the design philosophy of PNG. Chunk
registration is not automatic, although it is the intent of the
Registration Authority that it be straightforward when a new
chunk of potentially wide application is needed. The creation of
new critical chunk types is discouraged unless absolutely
necessary.</p>

<h2><a name="14Ordering">14.2 Behaviour of PNG editors</a></h2>

<p>A "PNG editor" is defined as a program that reads a PNG
datastream, makes modifications, and writes a new PNG datastream
while preserving as much ancillary information as possible. Two
examples of PNG editors are a program that adds or modifies text
chunks, and a program that adds a suggested palette to a
truecolour PNG datastream. Ordinary image editors are not PNG
editors because they usually discard all unrecognized information
while reading in an image.</p>

<p>To allow new chunk types to be added to PNG, it is necessary
to establish rules about the ordering requirements for all chunk
types. Otherwise a PNG editor does not know what to do when it
encounters an unknown chunk.</p>

<p>EXAMPLE Consider a hypothetical new ancillary chunk type that
is safe-to-copy and is required to appear after <a href=
"#11PLTE"><span class="chunk">PLTE</span></a> if <a href=
"#11PLTE"><span class="chunk">PLTE</span></a> is present. If a
program attempts to add a <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk and does not recognize the new
chunk, it may insert the <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> chunk in the wrong place, namely after
the new chunk. Such problems could be prevented by requiring PNG
editors to discard all unknown chunks, but that is a very
unattractive solution. Instead, PNG requires ancillary chunks not
to have ordering restrictions like this.</p>

<p>To prevent this type of problem while allowing for future
extension, constraints are placed on both the behaviour of PNG
editors and the allowed ordering requirements for chunks. The
safe-to-copy bit defines the proper handling of unrecognized
chunks in a datastream that is being modified.</p>

<!-- <ol start="1"> --><ol>
<li>If a chunk's safe-to-copy bit is 1, the chunk may be copied
to a modified PNG datastream whether or not the PNG editor
recognizes the chunk type, and regardless of the extent of the
datastream modifications.</li>

<li>If a chunk's safe-to-copy bit is 0, it indicates that the
chunk depends on the image data. If the program has made
<strong>any</strong> changes to <strong>critical</strong> chunks, including
addition, modification, deletion, or reordering of critical
chunks, then unrecognized unsafe chunks shall
<strong>not</strong> be copied to the output PNG datastream. (Of
course, if the program <strong>does</strong> recognize the chunk,
it can choose to output an appropriately modified version.)</li>

<li>A PNG editor is always allowed to copy all unrecognized
ancillary chunks if it has only added, deleted, modified, or
reordered <strong>ancillary</strong> chunks. This implies that it is not
permissible for ancillary chunks to depend on other ancillary
chunks.</li>

<li>PNG editors shall terminate on encountering an unrecognized
critical chunk type, because there is no way to be certain that a
valid datastream will result from modifying a datastream
containing such a chunk. (Simply discarding the chunk is not good
enough, because it might have unknown implications for the
interpretation of other chunks.) The safe/unsafe mechanism is
intended for use with ancillary chunks. The safe-to-copy bit will
always be 0 for critical chunks.</li>
</ol>

<p>The rules governing ordering of chunks are as follows.</p>

<!-- <ol start="5"> --><ol>
<li>When copying an unknown <strong>unsafe-to-copy</strong> ancillary
chunk, a PNG editor shall not move the chunk relative to any
critical chunk. It may relocate the chunk freely relative to
other ancillary chunks that occur between the same pair of
critical chunks. (This is well defined since the editor shall not
add, delete, modify, or reorder critical chunks if it is
preserving unknown unsafe-to-copy chunks.)</li>

<li>When copying an unknown <strong>safe-to-copy</strong> ancillary
chunk, a PNG editor shall not move the chunk from before <a href=
"#11IDAT"><span class="chunk">IDAT</span></a> to after <a href=
"#11IDAT"><span class="chunk">IDAT</span></a> or vice versa.
(This is well defined because <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> is always present.) Any other reordering
is permitted.</li>

<li>When copying a <strong>known</strong> ancillary chunk type, an editor
need only honour the specific chunk ordering rules that exist for
that chunk type. However, it may always choose to apply the above
general rules instead.</li>
</ol>

<p>These rules are expressed in terms of copying chunks from an
input datastream to an output datastream, but they apply in the
obvious way if a PNG datastream is modified in place.</p>

<p>See also 5.4: <a href="#5Chunk-naming-conventions"><span
class="xref">Chunk naming conventions</span></a>.</p>

<p>PNG editors that do not change the image data should not
change the <a href="#11tIME"><span class="chunk">tIME</span></a>
chunk. The Creation Time keyword in the <a href="#11tEXt"><span
class="chunk">tEXt</span></a>, <a href="#11zTXt"><span class=
"chunk">zTXt</span></a>, and <a href="#11iTXt"><span class=
"chunk">iTXt</span></a> chunks may be used for a user-supplied
time.</p>

<h2><a name="14Ordering-of-chunks">14.3 Ordering of
chunks</a></h2>

<h3><a name="14Ordering-of-critical-chunks">14.3.1 Ordering of
critical chunks</a></h3>

<p>Critical chunks may have arbitrary ordering requirements,
because PNG editors are required to terminate if they encounter
unknown critical chunks. For example <a href="#11IHDR"><span
class="chunk">IHDR</span></a> has the specific ordering rule that
it shall always appear first. A PNG editor, or indeed any
PNG-writing program, shall know and follow the ordering rules for
any critical chunk type that it can generate.</p>

<h3><a name="14Ordering-of-ancillary-chunks">14.3.2 Ordering of
ancillary chunks</a></h3>

<p>The strictest ordering rules for an ancillary chunk type
are:</p>

<!-- <ol start="1"> --><ol>
<li>Unsafe-to-copy chunks may have ordering requirements relative
to critical chunks.</li>

<li>Safe-to-copy chunks may have ordering requirements relative
to <a href="#11IDAT"><span class="chunk">IDAT</span></a>.</li>
</ol>

<p>The actual ordering rules for any particular ancillary chunk
type may be weaker. See for example the ordering rules for the
standard ancillary chunk types in 5.6: <a href=
"#5ChunkOrdering"><span class="xref">Chunk
ordering</span></a>.</p>

<p>Decoders shall not assume more about the positioning of any
ancillary chunk than is specified by the chunk ordering rules. In
particular, it is never valid to assume that a specific ancillary
chunk type occurs with any particular positioning relative to
other ancillary chunks.</p>

<p>EXAMPLE It is unsafe to assume that a particular private
ancillary chunk occurs immediately before <a href="#11IEND"><span
class="chunk">IEND</span></a>. Even if it is always written in
that position by a particular application, a PNG editor might
have inserted some other ancillary chunk after it. But it is safe
to assume that the chunk will remain somewhere between <a href=
"#11IDAT"><span class="chunk">IDAT</span></a> and <a href=
"#11IEND"><span class="chunk">IEND</span></a>.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1><a name="15Conformance">15 Conformance</a></h1>

<h2><a name="15ConfIntro">15.1 Introduction</a></h2>

<h3><a name="15ConfObjectives">15.1.1 Objectives</a></h3>

<p>This clause addresses conformance of PNG datastreams, PNG
encoders, PNG decoders, and PNG editors.</p>

<p>The primary objectives of the specifications in this clause
are:</p>

<!-- <ol start="1"> --><ol>
<li>to promote interoperability by eliminating arbitrary subsets
of, or extensions to, this International Standard;</li>

<li>to promote uniformity in the development of conformance
tests;</li>

<li>to promote consistent results across PNG encoders, decoders,
and editors;</li>

<li>to facilitate automated test generation.</li>
</ol>

<h3><a name="15ConfScope">15.1.2 Scope</a></h3>

<p>Conformance is defined for PNG datastreams and for PNG
encoders, decoders, and editors.</p>

<p>This clause addresses the PNG datastream and implementation
requirements including the range of allowable differences for PNG
encoders, PNG decoders, and PNG editors. This clause does not
directly address the environmental, performance, or resource
requirements of the encoder, decoder, or editor.</p>

<p>The scope of this clause is limited to rules for the open
interchange of PNG datastreams.</p>

<h2><a name="15ConformanceConf">15.2 Conformance conditions</a></h2>

<h3><a name="15FileConformance">15.2.1 Conformance of PNG
datastreams</a></h3>
<p>A PNG datastream conforms to this International Standard if
the following conditions are met.</p>
<ol>
<li>The PNG datastream contains a PNG signature as the first
content (see 5.2: <a href="#5PNG-file-signature"><span class=
"xref">PNG file signature</span></a>).</li>

<li>With respect to the chunk types defined in this International
Standard: 

<ul>
<li>the PNG datastream contains as its first chunk, an <a href=
"#11IHDR"><span class="chunk">IHDR</span></a> chunk, immediately
following the PNG signature;</li>

<li>the PNG datastream contains as its last chunk, an <a href=
"#11IEND"><span class="chunk">IEND</span></a> chunk.</li>
</ul>
</li>

<li>No chunks or other content follow the <a href="#11IEND"><span
class="chunk">IEND</span></a> chunk.</li>

<li>All chunks contained therein match the specification of the
corresponding chunk types of this International Standard. 
The PNG datastream shall obey the relationships among chunk types
defined in this International Standard.</li>

<li>The sequence of chunks in the PNG datastream obeys the
ordering relationship specified in this International
Standard.</li>

<li>All field values in the PNG datastream obey the relationships
specified in this International Standard producing the structure
specified in this International Standard.</li>

<li>No chunks appear in the PNG datastream other than those
specified in this International Standard or those defined
according to the rules for creating new chunk types as defined in
this International Standard.</li>

<li>The PNG datastream is encoded according to the rules of this
International Standard.</li>
</ol>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h3><a name="15ConformanceEncoder">15.2.2 Conformance of PNG
encoders</a></h3>

<p>A PNG encoder conforms to this International Standard if it
satisfies the following conditions.</p>

<!-- <ol start="1"> --><ol>
<li>All PNG datastreams that are generated by the PNG encoder are
conforming PNG datastreams.</li>

<li>When encoding input samples that have a sample depth that
cannot be directly represented in PNG, the encoder scales the
samples up to the next higher sample depth that is allowed by
PNG. The data are scaled in such a way that the high-order bits
match the original data.</li>

<li>Numbers greater than 127 are used when encoding experimental
or private definitions of values for any of the method or type
fields.</li>
</ol>

<h3><a name="15ConformanceDecoder">15.2.3 Conformance of PNG
decoders</a></h3>

<p>A PNG decoder conforms to this International Standard if it
satisfies the following conditions.</p>

<!-- <ol start="1"> --><ol>
<li>It is able to read any PNG datastream that conforms to this
International Standard, including both public and private chunks
whose types may not be recognized.</li>

<li>It supports all the standardized critical chunks, and all the
standardized compression, filter, and interlace methods and types
in any PNG datastream that conforms to this International
Standard.</li>

<li>Unknown chunk types are handled as described in <a href=
"#5Chunk-naming-conventions"><span class="xref">5.4 Chunk naming
conventions</span></a>. An unknown chunk type is <strong>not</strong>
treated as an error unless it is a critical chunk.</li>

<li>Unexpected values in fields of known chunks (for example, an
unexpected compression method in the <a href="#11IHDR"><span
class="chunk">IHDR</span></a> chunk) are treated as errors.</li>

<li>All types of PNG images (indexed-colour, truecolour,
greyscale, truecolour with alpha, and greyscale with alpha) are
processed. For example, decoders which are part of viewers
running on indexed-colour display hardware shall reduce
truecolour images to indexed format for viewing.</li>

<li>Encountering an unknown chunk in which the ancillary bit is 0
generates an error if the decoder is attempting to extract the
image.</li>

<li>A chunk type in which the reserved bit is set is treated as
an unknown chunk type.</li>

<li>All valid combinations of bit depth and colour type as
defined in 11.2.2: <a href="#11IHDR"><span class="xref"><span
class="chunk">IHDR</span> Image header</span></a> are
supported.</li>

<li>An error is reported if an unrecognized value is encountered
in the bit depth, colour type, compression method, filter method,
or interlace method bytes of the <a href="#11IHDR"><span class=
"chunk">IHDR</span></a> chunk.</li>

<li>When processing 16-bit greyscale or truecolour data in the <a
href="#11tRNS"><span class="chunk">tRNS</span></a> chunk, both
bytes of the sample values are evaluated to determine whether a
pixel is transparent.</li>

<li>When processing an image compressed by compression method 0,
the decoder assumes no more than that the complete image data is
represented by a single compressed datastream that is stored in
some number of <a href="#11IDAT"><span class=
"chunk">IDAT</span></a> chunks.</li>

<li>No assumptions are made concerning the positioning of any
ancillary chunk other than those that are specified by the chunk
ordering rules.</li>
</ol>

<h3><a name="15ConformanceEditor">15.2.4 Conformance of PNG
editors</a></h3>

<p>A PNG editor conforms to this International Standard if it satisfies the following conditions.</p>

<ol>
<li>It conforms to the requirements for PNG encoders.</li>

<li>It conforms to the requirements for PNG decoders.</li>

<li>It is able to encode all chunks that it decodes.</li>

<li>It preserves the ordering of the chunks presented within the
rules in 5.6: <a href="#5ChunkOrdering"><span class="xref">Chunk
ordering</span></a>.</li>

<li>It properly processes the safe-to-copy bit information and
preserves unknown chunks when the safe-to-copy rules permit
it.</li>

<li>Unless the user specifically permits lossy operations or the
editor issues a warning, it preserves all information required to
reconstruct the reference image exactly, except that the sample
depth of the alpha channel need not be preserved if it contains
only zero and maximum values. Operations such as changing the
colour type or rearranging the palette in an indexed-colour
datastream are permitted provided that the new datastream
losslessly represents the same reference image.</li>
</ol>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1 class="Annex"><a name="A-Conventions">Annex A</a></h1>

<p class="Annex">(informative)</p>

<h1 id="filemedia" class="Annex">File conventions and Internet media type</h1>

<h2><a name="A-File-name-extension">A.1 File name
extension</a></h2>

<p>On systems where file names customarily include an extension
signifying file type, the extension "<tt>.png</tt>" is
recommended for PNG files. Lower case "<tt>.png</tt>" is
preferred if file names are case-sensitive.</p>

<h2><a name="A-Media-type">A.2 Internet media type</a></h2>

<p>The internet media type "<tt>image/png</tt>" is the Internet
Media Type for PNG <a href="#2-RFC-2045"><span class=
"NormRef">[RFC-2045]</span></a>, <a href="#2-RFC-2048"><span
class="NormRef">[RFC-2048]</span></a>. It is recommended that
implementations also recognize the media type
"<tt>image/x-png</tt>".</p>

<h2><a name="A-Macintosh-file-layout">A.3 Macintosh file
layout</a></h2>

<p>In the Apple Computer Inc. Macintosh system, the following
conventions are recommended.</p>

<ol>
<li>The four-byte file type code for PNG files is
"<tt>PNGf</tt>". (This code has been registered with Apple
Computer Inc. for PNG files.) The creator code will vary
depending on the creating application.</li>

<li>The contents of the data fork is a PNG file exactly as
described in this International Standard.</li>

<li>The contents of the resource fork are unspecified. It may be
empty or may contain application-dependent resources.</li>

<li>When transferring a Macintosh PNG file to a non-Macintosh
system, only the data fork should be transferred.</li>
</ol>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1 class="Annex"><a name="B-NewChunksAppendix">Annex B</a></h1>

<p class="Annex">(informative)</p>

<h1 id="newchunks" class="Annex">Guidelines for new chunk types</h1>

<p>This International Standard allows extension through the
addition of new chunk types and new interlace, filter, and
compression methods. Such extensions might be made to the
standard either for experimental purposes or by organizations for
internal use.</p>

<p>Chunk types that are intended for general public use, or are
required for specific application domains, should be standardized
through registration (see 4.9 <a href=
"#4Concepts.Registration"><span class="xref">Extension and
registration</span></a>). The process for registration is defined
by the Registration Authority. The conventions for naming chunks
are given in 5.4: <a href="#5Chunk-naming-conventions"><span
class="xref">Chunk naming conventions</span></a>.</p>

<p>Some guidelines for defining private chunks are given
below.</p>

<!-- <ol start="1"> --><ol>
<li>Do not define new chunks that redefine the meaning of
existing chunks or change the interpretation of an existing
standardized chunk, e.g., do not add a new chunk to say that RGB
and alpha values actually mean CMYK.</li>

<li>Minimize the use of private chunks to aid portability.</li>

<li>Avoid defining chunks that depend on total datastream
contents. If such chunks have to be defined, make them critical
chunks.</li>

<li>For textual information that is representable in Latin-1
avoid defining a new chunk type. Use a <a href="#11tEXt"><span
class="chunk">tEXt</span></a> or <a href="#11zTXt"><span class=
"chunk">zTXt</span></a> chunk with a suitable keyword to identify
the type of information. For textual information that is not
representable in Latin-1 but which can be represented in UTF-8,
use an <a href="#11iTXt"><span class="chunk">iTXt</span></a>
chunk with a suitable keyword.</li>

<li>Group mutually dependent ancillary information into a single
chunk. This avoids the need to introduce chunk ordering
relationships.</li>

<li>Avoid defining private critical chunks.</li>
</ol>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1 class="Annex"><a name="C-GammaAppendix">Annex C</a></h1>

<p class="Annex">(informative)</p>

<h1 id="gammachromaticity" class="Annex">Gamma and chromaticity</h1>

<p>Gamma is a numerical parameter used to describe approximations
to certain non-linear transfer functions encountered in image
capture and reproduction. Gamma is the exponent in a power law
function. For example the function:</p>

<p><tt>intensity = (voltage +
constant)<sup>exponent</sup></tt></p>

<p>which is used to model the non-linearity of cathode ray tube
(CRT) displays. It is often assumed, as in this International
Standard, that the constant is zero.</p>

<p>For the purposes of this International Standard, it is
convenient to consider five places in a general image pipeline at
which non-linear transfer functions may occur and which may be
modelled by power laws. The characteristic exponent associated
with each is given a specific name.</p>

<table class="Regular" summary=
"This table describes characteristic exponents">
<tr>
<td class="Regular"><tt>input_exponent</tt> </td>
<td class="Regular">the exponent of the image sensor.</td>
</tr>

<tr>
<td class="Regular"><tt>encoding_exponent</tt> </td>
<td class="Regular">the exponent of any transfer function performed by the
process or device writing the datastream.</td>
</tr>

<tr>
<td class="Regular"><tt>decoding_exponent</tt> </td>
<td class="Regular">the exponent of any transfer function performed by the
software reading the image datastream.</td>
</tr>

<tr>
<td class="Regular"><tt>LUT_exponent</tt> </td>
<td class="Regular">the exponent of the transfer function applied between the
frame buffer and the display device (typically this is applied by
a Look Up Table).</td>
</tr>

<tr>
<td class="Regular"><tt>output_exponent</tt> </td>
<td class="Regular">the exponent of the display device. For a CRT, this is
typically a value close to 2.2.</td>
</tr>
</table>

<p>It is convenient to define some additional entities that
describe some composite transfer functions, or combinations of
stages.</p>

<table class="Regular" summary=
"This table characterises additional entities that are used to describe transfer functions">
<tr>
<td class="Regular"><tt>display_exponent</tt> </td>
<td class="Regular">exponent of the transfer function applied between the frame
buffer and the display surface of the display device.<br class="xhtml" />
<tt>display_exponent = LUT_exponent * output_exponent</tt> </td>
</tr>

<tr>
<td class="Regular"><tt>gamma</tt> </td>
<td class="Regular">exponent of the function mapping display output intensity to
samples in the PNG datastream.<br class="xhtml" />
<tt>gamma = 1.0 / (decoding_exponent * display_exponent)</tt>
</td>
</tr>

<tr>
<td class="Regular"><tt>end_to_end_exponent</tt> </td>
<td class="Regular">the exponent of the function mapping image sensor input
intensity to display output intensity. This is generally a value
in the range 1.0 to 1.5.</td>
</tr>
</table>

<p>The PNG <a href="#11gAMA"><span class="chunk">gAMA</span></a>
chunk is used to record the gamma value. This information may be
used by decoders together with additional information about the
display environment in order to achieve, or approximate, the
desired display output.</p>

<p>Additional information about this subject may be found in the
references <a href="#G-GAMMA-TUTORIAL"><span class=
"bibref">[GAMMA-TUTORIAL]</span></a>, <a href=
"#G-GAMMA-FAQ"><span class="bibref">[GAMMA-FAQ]</span></a>, and
<a href="#G-POYNTON"><span class="bibref">[POYNTON]</span></a>
(especially chapter 6).</p>

<p>Background information about chromaticity and colour spaces
may be found in references <a href="#G-COLOUR-TUTORIAL"><span
class="bibref">[COLOUR-TUTORIAL]</span></a>, <a href=
"#G-COLOUR-FAQ"><span class="bibref">[COLOUR-FAQ]</span></a>, <a
href="#G-HALL"><span class="bibref">[HALL]</span></a>, <a href=
"#G-KASSON"><span class="bibref">[KASSON]</span></a>, <a href=
"#G-LILLEY"><span class="bibref">[LILLEY]</span></a>, <a href=
"#G-STONE"><span class="bibref">[STONE]</span></a>, and <a href=
"#G-TRAVIS"><span class="bibref">[TRAVIS]</span></a>.</p>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1 class="Annex"><a name="D-CRCAppendix">Annex D</a></h1>

<p class="Annex">(informative)</p>

<h1 id="samplecrc" class="Annex">Sample Cyclic Redundancy Code
implementation</h1>

<p>The following sample code represents a practical
implementation of the CRC (Cyclic Redundancy Check) employed in
PNG chunks. (See also ISO 3309 <a href="#2-ISO-3309"><span class=
"NormRef">[ISO-3309]</span></a> or ITU-T V.42 <a href=
"#G-ITU-T-V42"><span class="bibref">[ITU-T-V42]</span></a> for a
formal specification.)</p>

<p>The sample code is in the ISO C <a href="#2-ISO-9899"><span
class="NormRef">[ISO-9899]</span></a> programming language. The
hints in <a href="#D-tabled1"><span class="tabref">Table
D.1</span></a> may help non-C users to read the code more
easily.</p>

<table class="Regular" summary=
"This table gives hints for reading the CRC code">
<caption><a name="D-tabled1"><b>Table D.1 &mdash; Hints for
reading ISO C code</b></a></caption>

<tr>
<td class="Regular"><tt>&amp;</tt> </td>
<td class="Regular">Bitwise AND operator.</td>
</tr>

<tr>
<td class="Regular"><tt>^</tt> </td>
<td class="Regular">Bitwise exclusive-OR operator.</td>
</tr>

<tr>
<td class="Regular"><tt>&gt;&gt;</tt> </td>
<td class="Regular">Bitwise right shift operator. When applied to an unsigned
quantity, as here, right shift inserts zeroes at the left.</td>
</tr>

<tr>
<td class="Regular"><tt>!</tt> </td>
<td class="Regular">Logical NOT operator.</td>
</tr>

<tr>
<td class="Regular"><tt>++</tt> </td>
<td class="Regular">"<tt>n++</tt>" increments the variable <tt>n</tt>. In "for"
loops, it is applied after the variable is tested.</td>
</tr>

<tr>
<td class="Regular"><tt>0xNNN</tt> </td>
<td class="Regular"><tt>0x</tt> introduces a hexadecimal (base 16) constant.
Suffix <tt>L</tt> indicates a long value (at least 32 bits).</td>
</tr>
</table>

<hr class="xhtml" />
<pre>
   /* Table of CRCs of all 8-bit messages. */
   unsigned long crc_table[256];
   
   /* Flag: has the table been computed? Initially false. */
   int crc_table_computed = 0;
   
   /* Make the table for a fast CRC. */
   void make_crc_table(void)
   {
     unsigned long c;
     int n, k;
   
     for (n = 0; n &lt; 256; n++) {
       c = (unsigned long) n;
       for (k = 0; k &lt; 8; k++) {
         if (c &amp; 1)
           c = 0xedb88320L ^ (c &gt;&gt; 1);
         else
           c = c &gt;&gt; 1;
       }
       crc_table[n] = c;
     }
     crc_table_computed = 1;
   }
  
</pre>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<pre>
   /* Update a running CRC with the bytes buf[0..len-1]--the CRC
      should be initialized to all 1's, and the transmitted value
      is the 1's complement of the final running CRC (see the
      crc() routine below). */
   
   unsigned long update_crc(unsigned long crc, unsigned char *buf,
                            int len)
   {
     unsigned long c = crc;
     int n;
   
     if (!crc_table_computed)
       make_crc_table();
     for (n = 0; n &lt; len; n++) {
       c = crc_table[(c ^ buf[n]) &amp; 0xff] ^ (c &gt;&gt; 8);
     }
     return c;
   }
   
   /* Return the CRC of the bytes buf[0..len-1]. */
   unsigned long crc(unsigned char *buf, int len)
   {
     return update_crc(0xffffffffL, buf, len) ^ 0xffffffffL;
   }
</pre>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1 class="Annex"><a name="E-Resources">Annex E</a></h1>

<p class="Annex">(informative)</p>

<h1 id="onlineresources" class="Annex">Online resources</h1>

<h2><a name="E-Intro">Introduction</a></h2>

<p>This annex gives the locations of some Internet resources for
PNG software developers. By the nature of the Internet, the list
is incomplete and subject to change.</p>

<h2><a name="E-Archive-sites">Archive sites</a></h2>

<p>This International Standard can be found at
<a href="http://www.w3.org/TR/2003/REC-PNG-20031110/index.html"
><code>http://www.w3.org/TR/2003/REC-PNG-20031110/index.html</code></a>.</p>

<h2><a name="E-icc-profile-specs">ICC profile
specifications</a></h2>

<p>ICC profile specifications are available at: <a href=
"http://www.color.org/"><code>http://www.color.org/</code></a></p>

<h2><a name="E-PNG-home-page">PNG web site</a></h2>

<p>There is a World Wide Web site for PNG at <a href=
"http://www.libpng.org/pub/png/"><code>http://www.libpng.org/pub/png/</code></a>.
This page is a central location for current information about PNG
and PNG-related tools.</p>

<p>Additional documentation and portable C code for deflate,
inflate, and an optimized implementation of the CRC algorithm are
available from the zlib web site,
<a href=
"http://www.zlib.org/"><code>http://www.zlib.org/</code></a>.</p>

<h2><a name="E-Sample-implementation">Sample implementation and
test images</a></h2>

<p>A sample implementation in portable C, <strong>libpng</strong>, is
available at <a href=
"http://www.libpng.org/pub/png/libpng.html"><code>http://www.libpng.org/pub/png/libpng.html</code></a>.
Sample viewer and encoder applications of libpng are available at
<a href=
"http://www.libpng.org/pub/png/book/sources.html"><code>http://www.libpng.org/pub/png/book/sources.html</code></a>
and are described in detail in <i>PNG: The Definitive Guide</i>
<a href="#G-ROELOFS">[ROELOFS]</a>. Test images can also be
accessed from the PNG web site.</p>

<h2><a name="E-Email">Electronic mail</a></h2>

<p>Queries concerning PNG developments may be addressed to <a href=
"mailto:png-group@w3.org"><tt>png-group@w3.org</tt></a>. 
<!-- ************Page Break******************* -->
</p>

<!-- ************Page Break******************* -->
<h1 class="Annex"><a name="F-Relationship">Annex F</a></h1>

<p class="Annex">(informative)</p>

<h1 id="relationshiptofirstedition" class="Annex">Relationship to W3C PNG</h1>

<p>This International Standard is strongly based on W3C
Recommendation PNG Specification Version 1.0 <a href=
"#G-PNG-1.0">[PNG-1.0]</a> which was reviewed by W3C members,
approved as a W3C Recommendation, and published in October 1996
according to the established W3C process. Subsequent amendments
to the PNG Specification have also been incorporated into this
International Standard <a href="#G-PNG-1.0">[PNG-1.1]</a>, <a
href="#G-PNG-1.0">[PNG-1.2]</a>.</p>

<p>A complete review of the document has been done by ISO/IEC/JTC
1/SC 24 in collaboration with W3C in order to transform this
recommendation into an ISO/IEC international standard. A major
design goal during this review was to avoid changes that will
invalidate existing files, editors, or viewers that conform to
W3C Recommendation PNG Specification Version 1.0.</p>

<p>The W3C PNG Recommendation was developed with major
contribution from the following people.</p>

<h2><a name="F-Editor10">Editor (Version 1.0)</a></h2>

<p>Thomas Boutell, <span class="email">boutell @ boutell.com</span></p>

<h2><a name="F-Editor12">Editor (Versions 1.1 and 1.2)</a></h2>

<p>Glenn Randers-Pehrson, <span class="email">randeg @ alum.rpi.edu</span></p>

<h2><a name="F-ContribEditor10">Contributing Editor (Version
1.0)</a></h2>

<p>Tom Lane, <span class="email">tgl @ sss.pgh.pa.us</span></p>

<h2><a name="F-ContribEditor12">Contributing Editor (Versions 1.1
and 1.2)</a></h2>

<p>Adam M. Costello, <span class="email">png-spec.amc @ nicemice.net</span></p>

<h2><a name="F-Authors">Authors (Versions 1.0, 1.1, and 1.2
combined)</a></h2>

<p><strong>Authors' names are presented in alphabetical
order.</strong></p>

<ul>
<li><a href="http://www.alumni.caltech.edu/~madler/">Mark Adler</a>,
<span class="email">madler @ alumni.caltech.edu</span></li>

<li><a href="http://www.boutell.com/boutell/">Thomas Boutell</a>,
<span class="email">boutell @ boutell.com</span></li>

<li>John Bowler, <span class="mail">jbowler @ acm.org</span></li>

<li><a href="http://www.df.lth.se/~cb/">Christian Brunschen</a>,
<span class="email">cb @ brunschen.com</span></li>

<li><a href="http://www.nicemice.net/amc/">Adam M.
Costello</a>, <span class=
"email">png-spec.amc @ nicemice.net</span></li>

<li><a href="http://www.piclab.com/">Lee Daniel Crocker</a>,
<span class="email">lee @ piclab.com</span></li>

<li><a href=
"http://www-mddsp.enel.ucalgary.ca/People/adilger/">Andreas
Dilger</a>, <span class=
"email">adilger @ turbolabs.com</span></li>

<li><a href="http://www.fromme.com/">Oliver Fromme</a>, <span
class="email">oliver @ fromme.com</span></li>

<li><a href="http://www.teaser.fr/~jlgailly/">Jean-loup
Gailly</a>, <span class="email">jloup @ gzip.org</span></li>

<li>Chris Herborth, <span class=
"email">chrish @ pobox.com</span></li>

<li>Alex Jakulin, <span class=
"email">jakulin @ acm.org</span></li>

<li>Neal Kettler, <span class=
"email">neal @ westwood.com</span></li>

<li>Tom Lane, <span class="email">tgl @ sss.pgh.pa.us</span></li>

<li>Alexander Lehmann, <span class=
"email">lehmann @ usa.net</span></li>


<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->

<li><a href="http://www.w3.org/People/chris/">Chris Lilley</a>,
<span class="email">chris @ w3.org</span></li>

<li>Dave Martindale, <span class=
"email">davem @ cs.ubc.ca</span></li>

<li>Owen Mortensen, <span class="email">ojm @ acm.org</span></li>

<li>Keith S. Pickens, <span class=
"email">ksp @ rice.edu</span></li>

<li><a href="http://www.users.qwest.net/~lionlad/">Robert P. Poole</a>, <span class=
"email">lionlad @ qwest.net</span></li>

<li>Glenn Randers-Pehrson, <span class=
"email">randeg @ alum.rpi.edu</span></li>

<li><a href="http://pobox.com/~newt/">Greg Roelofs</a>, <span
class="email">newt @ pobox.com</span></li>

<li><a href="http://www.schaik.com/">Willem van Schaik</a>, <span
class="email">willem @ schaik.com</span></li>

<li>Guy Schalnat, <span class=
"email">gschal @ infinet.com</span></li>

<li>Paul Schmidt, <span class=
"email">pschmidt @ photodex.com</span></li>

<li>Michael Stokes, <span class=
"email">mistokes @ microsoft.com</span></li>

<li>Tim Wegner, <span class=
"email">twegner @ phoenix.net</span></li>

<li>Jeremy Wohl, <span class=
"email">jeremyw @ evantide.com</span></li>
</ul>

<h2><a name="F-ChangeList">List of changes between W3C
Recommendation PNG Specification Version 1.0 and this
International Standard</a></h2>

<h3><a name="F-EditorialChanges">Editorial changes</a></h3>

<p>The document has been reformatted according to the
requirements of ISO.</p>

<!-- <ol start="1"> --><ol>
<li>A concepts clause has been introduced.</li>

<li>Conformance for datastreams, encoders, decoders, and editors
has been defined in a conformance clause.</li>
</ol>

<h3><a name="F-TechnicalChanges">Technical changes</a></h3>

<!-- <ol start="1"> --><ol>
<li>New chunk types introduced in PNG version 1.1 and 1.2 have
been incorporated (<a href="#11iCCP"><span class=
"chunk">iCCP</span></a>, <a href="#11iTXt"><span class=
"chunk">iTXt</span></a>, <a href="#11sRGB"><span class=
"chunk">sRGB</span></a>, <a href="#11sPLT"><span class=
"chunk">sPLT</span></a>).
In the
<a href="#11iTXt"><span class=
"chunk">iTXt</span></a>
chunk, the language tag has been updated from RFC 1766 to RFC 3066.</li>

<li>In accord with version 1.1, the scope of the 31-bit limit on
chunk lengths and image dimensions has been extended to apply to
all four-byte unsigned integers. The value -2<sup>31</sup> is not
allowed in signed integers.</li>

<li>The redefinition of <a href="#11gAMA"><span class=
"chunk">gAMA</span></a> to be in terms of the desired display
output rather than the original scene, introduced in PNG version
1.1, has been incorporated.</li>

<li>The use of the <a href="#11PLTE"><span class=
"chunk">PLTE</span></a> and <a href="#11hIST"><span class=
"chunk">hIST</span></a> chunks in non-indexed-colour images has
been discouraged in favour of the <a href="#11sPLT"><span class=
"chunk">sPLT</span></a> chunk.</li>

<li>Some recommendations for PNG encoders, decoders, and editors
have been strengthened to requirements. These changes do not
affect the conformance of PNG datastreams, and do not compromise
interoperability.</li>

<li>The sample depth of channels not mentioned in the <a href=
"#11sBIT"><span class="chunk">sBIT</span></a> chunk has been
clarified.</li>
</ol>

<!-- ************Page Break******************* -->
<!-- ************Page Break******************* -->
<h1 class="Annex"><a name="G-References">Bibliography</a></h1>

<dl>
<dt><a name="G-COLOUR-FAQ">[COLOUR-FAQ]</a></dt>

<dd>Poynton, C., "Colour FAQ".<br class="xhtml" />
 <a href=
"http://www.poynton.com/ColorFAQ.html">
<code>http://www.poynton.com/ColorFAQ.html</code></a></dd>

<dt><a name="G-COLOUR-TUTORIAL">[COLOUR-TUTORIAL]</a></dt>

<dd>PNG Group, "Colour tutorial".<br class="xhtml" />
 <a href=
"http://www.libpng.org/pub/png/spec/1.2/PNG-ColorAppendix.html"><code>
http://www.libpng.org/pub/png/spec/1.2/PNG-ColorAppendix.html</code></a></dd>

<dt><a name="G-GAMMA-TUTORIAL">[GAMMA-TUTORIAL]</a></dt>

<dd>PNG Group, "Gamma tutorial".<br class="xhtml" />
 <a href=
"http://www.libpng.org/pub/png/spec/1.2/PNG-GammaAppendix.html"><code>
http://www.libpng.org/pub/png/spec/1.2/PNG-GammaAppendix.html</code></a></dd>

<dt><a name="G-GAMMA-FAQ">[GAMMA-FAQ]</a></dt>

<dd>Poynton, C., "Gamma FAQ".<br class="xhtml" />
 <a href=
"http://www.poynton.com/Poynton-color.html">
<code>http://www.poynton.com/Poynton-color.html</code></a></dd>

<dt><a name="G-HALL">[HALL]</a></dt>

<dd>Hall, Roy, <i>Illumination and Color in Computer Generated
Imagery</i>. Springer-Verlag, New York, 1989. ISBN
0-387-96774-5.</dd>

<dt><a name="G-ICC">[ICC]</a></dt>

<dd>The International Color Consortium.<br class="xhtml" />
 <a href=
"http://www.color.org/"><code>http://www.color.org/</code></a></dd>

<dt><a name="G-ISO-3664">[ISO-3664]</a></dt>

<dd>ISO 3664:2000, <i>Viewing conditions &mdash; Graphic
technology and photography</i>.</dd>

<dt><a name="G-ITU-R-BT709">[ITU-R-BT709]</a></dt>

<dd>International Telecommunications Union, <i>Basic Parameter
Values for the HDTV Standard for the Studio and for International
Programme Exchange</i>, ITU-R Recommendation BT.709 (formerly CCIR
Rec. 709), 1990.</dd>

<dt><a name="G-ITU-T-V42">[ITU-T-V42]</a></dt>

<dd>International Telecommunications Union, <i>Error-correcting
Procedures for DCEs Using Asynchronous-to-Synchronous
Conversion</i>, ITU-T Recommendation V.42, 1994, Rev. 1.</dd>

<dt><a name="G-KASSON">[KASSON]</a></dt>

<dd>Kasson, J., and W. Plouffe, "An Analysis of Selected Computer
Interchange Color Spaces", <i>ACM Transactions on Graphics</i>,
vol. 11, no. 4 , pp. 373-405, 1992.</dd>

<dt><a name="G-LILLEY">[LILLEY]</a></dt>

<dd>Lilley, C., F. Lin, W.T. Hewitt, and T.L.J. Howard, <i>Colour
in Computer Graphics</i>. CVCP, Sheffield, 1993. ISBN
1-85889-022-5.<br class="xhtml" />
<!-- Also available from<br class="xhtml" />
 <a href=
"http://www.man.ac.uk/MVC/training/gravigs/colour/"><code>http://www.man.ac.uk/MVC/training/gravigs/colour/</code></a>
--></dd>

<dt><a name="G-ROELOFS">[ROELOFS]</a></dt>

<dd>Roelofs, G., <i>PNG: The Definitive Guide</i>, O'Reilly &amp;
Associates Inc, Sebastopol, CA, 1999. ISBN 1-56592-542-4.
See also <a href="http://www.libpng.org/pub/png/pngbook.html">
<code>http://www.libpng.org/pub/png/pngbook.html</code>
</a></dd>

<dt><a name="G-PAETH">[PAETH]</a></dt>

<dd>Paeth, A.W., "Image File Compression Made Easy", in
<i>Graphics Gems II</i>, James Arvo, editor. Academic Press, San
Diego, 1991. ISBN 0-12-064480-0.</dd>

<dt><a name="G-PNG-1.0">[PNG-1.0]</a></dt>

<dd>W3C Recommendation, "PNG (Portable Network Graphics)
Specification, Version 1.0", 1996. Available in several formats
from<br class="xhtml" />
 <a href=
"http://www.w3.org/TR/REC-png-961001"><code>http://www.w3.org/TR/REC-png-961001</code></a>
and from<br class="xhtml" />
 <a href=
"http://www.libpng.org/pub/png/spec/1.0/"><code>http://www.libpng.org/pub/png/spec/1.0/</code></a></dd>

<dt><a name="G-PNG-1.1">[PNG-1.1]</a></dt>

<dd>PNG Development Group, "PNG (Portable Network Graphics)
Specification, Version 1.1", 1999. Available 
from<br class="xhtml" />
 <a href=
"http://www.libpng.org/pub/png/spec/1.1/"><code>http://www.libpng.org/pub/png/spec/1.1/</code></a></dd>

<dt><a name="G-PNG-1.2">[PNG-1.2]</a></dt>

<dd>PNG Development Group, "PNG (Portable Network Graphics)
Specification, Version 1.2", 1999. Available from<br class="xhtml" />
 <a href=
"http://www.libpng.org/pub/png/spec/1.2/"><code>http://www.libpng.org/pub/png/spec/1.2/</code></a></dd>

<dt><a name="G-PNG-EXTENSIONS">[PNG-REGISTER]</a></dt>

<dd>PNG Development Group, "Register of PNG Public Chunks and Keywords".
Available in several formats from:<br class="xhtml" />
<a href=
"http://www.libpng.org/pub/png/spec/register/"><code>http://www.libpng.org/pub/png/spec/register/</code></a></dd>

<dt><a name="G-POSTSCRIPT">[POSTSCRIPT]</a></dt>

<dd>Adobe Systems Incorporated, <i>PostScript Language Reference
Manual</i>, 2nd edition. Addison-Wesley, Reading, 1990. ISBN
0-201-18127-4.</dd>

<dt><a name="G-POYNTON">[POYNTON]</a></dt>

<dd>Poynton, Charles A., <i>A Technical Introduction to Digital
Video</i>. John Wiley and Sons, Inc., New York, 1996. ISBN
0-471-12253-X.</dd>

<dt><a name="G-SMPTE-170M">[SMPTE-170M]</a></dt>

<dd>Society of Motion Picture and Television Engineers,
<i>Television &mdash; Composite Analog Video Signal &mdash; NTSC
for Studio Applications</i>, SMPTE-170M, 1994.</dd>

<dt><a name="G-STONE">[STONE]</a></dt>

<dd>Stone, M.C., W.B. Cowan, and J.C. Beatty, "Color gamut
mapping and the printing of digital images", <i>ACM Transactions on
Graphics</i>, vol. 7, no. 3, pp. 249-292, 1988.</dd>

<dt><a name="G-TIFF-6.0">[TIFF-6.0]</a></dt>

<dd>TIFF<sup>TM</sup> Revision 6.0, Aldus Corporation, June
1992.</dd>

<dt><a name="G-TRAVIS">[TRAVIS]</a></dt>

<dd>Travis, David, <i>Effective Color Displays &mdash; Theory and
Practice</i>. Academic Press, London, 1991. ISBN
0-12-697690-2.</dd>

<dt><a name="G-ZL">[ZL]</a></dt>

<dd>J. Ziv and A. Lempel, "A Universal Algorithm for Sequential
Data Compression", <i>IEEE Transactions on Information
Theory</i>, vol. IT-23, no. 3, pp. 337 - 343, 1977.</dd>
</dl>

<p>Additional documentation and portable C code for deflate,
inflate, and an optimized implementation of the CRC algorithm are
available from the zlib web site,
<a href=
"http://www.zlib.org/"><code>http://www.zlib.org/</code></a>.</p>
</body>
</html>